The first anesthesia information systems collected only intra-operative data; in fact, most currently available commercial systems still concentrate on generating a computerized anesthesia (intra-operative) record. However, if the goal of implementing an anesthesia information system is to close the loop between what we do and how well the patients do, information beyond the intra-operative period must be collected. Of course the analysis of these data and the resulting information presents a challenge.
More Than an Extension of the Intra-Operative Record
The "post-anesthesia record" might typically be thought of as the Post-Anesthesia Care Unit (PACU) clinical documentation. Since the PACU is an extension of what happens in the operating room, similar components are required to make the system work; i.e., interfaces to physiological monitors and software implementations that achieve clinical documentation. However, to realize the promise of the benefit of implementing an anesthesia information system, it must include data elements required to fulfill regulatory/compliance requirements and to support administrative functions of both the responsible physicians and the hospital. Therefore, the Post-Anesthesia Record must be a component of an anesthesia information system that encompasses the horizontal and vertical scope of such a system. The Post-Anesthesia Record should include the anesthesiologist’s postoperative visit. Furthermore, because both the surgeon and the anesthesiologist are both interested in postoperative events, the clinical documentation and reports from these different specialties should be coordinated. The postoperative visit should also include data describing the patient’s satisfaction. Measuring patient satisfaction is a science unto itself. Unfortunately, there are few studies in the anesthesia literature to show how to do this properly. Anesthesia, surgery, and nursing services capture many of the same elements. How to reduce unnecessary duplication and how to share data calls for clever and intuitive implementations. For example, if the anesthesia History and Physical Examination include the fact that the patient is allergic to penicillin or has hypertension, these data elements should automatically populate the nursing documentation. Who is the authoritative source for such information? The system in use at the University of Michigan uses a "split-screen" and requires a confirmatory step that carries data elements forward from one clinician’s documentation to another.
Standardization of the Post-Anesthesia Record
We need standards that allow us to compare data from one practitioner to another or one institution to another. Some things are easy to define and quantitate (e.g. death); others much more difficult (e.g. nausea, pain). Patients are as concerned about nausea and vomiting as they are about pain and we do not have good ways to measure it. The APSF is actively pursuing the compilation of a data dictionary to facilitate data analysis and comparison with the ultimate goal of improving patient safety.
Post-Anesthesia Record as Part of a System
The information contained in the anesthesiology documentation, pre-, intra- and postoperatively must interface with an institutional clinical data repository or order entry system (See Table below).
Components of Post Operative Point-of-Care Clinical Documentation
|• PACU flow sheet|
|• Summary of Intra-operative Anesthesia Care|
|• Automated acquisition and display of real time vital signs and trends|
|• Record and display notes and annotations, including pain scores (the fifth vital sign)|
|• Medication administration documentation|
|• Clinical summary of post anesthesia care|
|• Confirmation of preoperatively determined postoperative care plan|
|• Explanation of potential QA indicators for events|
|• Pain care plan after discharge from PACU|
If a patient’s trachea was difficult to intubate, how does this vital airway information get back to the institution’s clinical data repository so it will be always available to others? The most basic requirement is a printout, but its distribution presents major hurdles, as we know from bitter experience with the traditional handwritten medical records. A web-based view makes the information readily available to many simultaneously at multiple locations. It raises issues of confidentiality, which are as relevant for the postoperative record as for any other phase of the patient’s medical experiences. The "data" associated with an encounter should ideally be supplied to other systems as codified data; not just a view of the information. This is very important for several reasons. One is documentation of comorbidities and complications. Hospitals take care of very sick patients. A principal role of the anesthesiologist is to identify and manage comorbidities. Since a large proportion of patients in a hospital undergo surgery, anesthesiologists can make an important contribution to documenting the severity of illness. At the same time, they want to be the recipient of relevant information in a seamlessly linked information system, as relevant for the postoperative period as at any other time during the other phases of anesthesia care.
Post-Anesthesia Record and Quality Assurance
Anesthesiologists may think of Quality Assurance data as "post-op data." However, QA may be applied to each phase of the perioperative process. We typically think of intra-operative and postoperative events, but what about QA on the preoperative evaluation or the patient’s stay in the PACU? Do patients go to the operating room that should have received further evaluation? Was the patient’s nausea or pain in the PACU treated in a timely and effective way?
Currently, the QA process is typically post hoc and the reports of events are generated voluntarily. With the implementation of computerized information systems, it makes sense to include logic that generates reports of "errors" as defined by the Institute of Medicine report; that is, actions that result in an outcome other than what was expected. Software logic may be applied to the accurate and complete clinical documentation and then suggest a determination whether a case involved an error. For example, if an ASA Physical Status One patient receives epinephrine, something happened, and further details would be invaluable to anyone reviewing this case. With the integration of physiological monitors, other events will be suggested, however, a consensus will need to be achieved within a department or group, and hopefully, within our specialty, about what constitutes an "event." If the oxygen saturation drops below 90% for greater than five minutes, is that hypoxemia? In the future, it is likely that real time prompts ("smart alarms") will suggest the presence of an adverse event or error and suggests additional documentation be provided to help the person who will eventually be reviewing that case. These comments must be protected from outside discovery by the QA process.
Many of the problems with modern health care are systems problems. As noted by Dr. Avedis Donabedian, a well-known health care quality researcher who spoke about his own medical care: "People have a big problem understanding the relationship between quality and systems. Many doctors seek refuge in the allegation that they are good clinicians but the system is wrong, without realizing that they are the key aspects of the system… The system is the problem… " (The New York Times, June 12, 2001) To make improvements, we must change our systems. Information systems provide powerful tools to reengineer systems and work processes. Since operating room suites provide much of the income for many health systems, improvements in perioperative systems can have a large impact on the institution. Realizing this, institutions are beginning to make the appropriate investments in information systems. As physicians, we need data that we can transform into information and apply as knowledge. The data elements we need to achieve these goals go beyond what is required to provide a complete and accurate clinical record but they are nonetheless important to achieve the potential benefit of clinical information systems. We must participate and insist on follow through. The so-often neglected Post-Anesthesia Record must be an integral part of such a system.
Dr. O’Reilly is from the Department of Anesthesiology, University of Michigan, Ann Arbor.