Pre-anesthesia evaluation data are an essential part of good anesthesia care. The pre-anesthesia evaluation is part of medical informatics as applied to anesthesia. The study of the statistical association between pre-anesthesia evaluation elements and untoward incidents during the intra- and postoperative periods requires data that are currently not available and sure need to be collected. The results from such studies will lead to improved patient care and will enable us to identify the patients who need it most and to focus additional testing and intense monitoring on specific groups of patients. The immediate anesthetic needs of clinicians can easily be merged with the needs of medical informatics to result in a pre-anesthesia evaluation process that supports excellent patient care while providing the necessary informatics data.
Importance of Pre-Anesthesia Evaluation Data
When anesthesiologists are asked what information they want to get from a database, the usual response is “How many patients [of some age range] with [some medical problem or problems] have [some abnormal condition] during [some anesthesia technique]?” For example, “How many patients of all ages with diabetes have bradycardia during spinal anesthesia?” During pre-anesthesia evaluation, the clinician collects and documents the presence or absence of “some medical problem or problems.” Through this evaluation, clinicians collect important information that provides answers to the categorical questions posed above. Without this information, you could only ask how many patients had bradycardia. You could not relate this vital sign incident to pre-existing medical conditions to assess predictive risk factors. The value of a database lies in its ability to give us information that predicts which patients are at risk, need more testing, and more intensive monitoring.
Most anesthesiologist/administrators who are charged with planning a pre-anesthesia evaluation process understand clinical needs but have not had training in medical informatics. A collection of standard data elements, using agreed-upon standard definitions in a way that can be easily stored and searched in a computer database is the basis of information retrieval. Informatics requires yes/no answers, or answers selected from a pre-defined list of choices. Blobs of text (free-form notes) are the anathema of medical informatics because text is not usable as data. Data cannot be queried from free-form text in a database. Invariably, clinicians use multiple words or abbreviations in text as synonyms. Humans can deal with this imprecision, but computers cannot. The needs of informatics inevitably point to a structured process.
Clinicians spend the time to collect pre-anesthesia data because they believe this effort makes a significant contribution to patient safety and better anesthesia care. When pre-anesthesia information is collected and stored using a computer-based system the information becomes promptly available 24 hours a day when it is needed for subsequent anesthetics. A computer-based system can alert clinicians to problems that occurred during past anesthetics as a part of the pre-anesthesia data review.
Clinicians today also face an intensive pressure to increase productivity by spending less time with each patient. Use of a structured technique, whether a form or direct entry into a computer program, can be used to diminish the time a clinician needs to spend with documentation. Some clinicians have formed an association between their professional standing and documenting their pre-anesthesia evaluation as free-text in the progress notes of the patient’s chart, following the custom of other (often primary care) medical practitioners. This emotional attachment stands in opposition to the adoption of structured processes that advance medical informatics. Anesthesia has blazed the trail of patient safety before, and as the value of data becomes increasingly apparent, anesthesia clinicians will again have the opportunity to set the path for other medical specialties to follow.
There is a convergence between the needs of medical informatics and the clinical needs of anesthesia practice. It is important to overcome the negative emotional aspects of using forms and other structured processes as we accomplish our clinical and information goals.
Standard Data Elements
The lack of a standard for data elements is a limit to the pre-anesthesia evaluationÑboth as a data source and as a medical process. The ASA’s Basic Standard for Pre-Anesthesia Care from 1987 must be updated to be sufficiently definitive for medical informatics. The first step in standardization is to build a standard minimum list of data elements (systems or disease conditions) to be collected, acceptable to the whole anesthesia community. Intentionally, the list must be kept simple: 10 Ð 20 systems or disease conditions. If the standard becomes too complicated, it will exacerbate the time pressure that troubles clinicians in day-to-day practice. Each element representing a system or the presence of a disease must be answerable by a yes or no answer, and evaluators must accept the need to make a yes or no decision as the evaluation is conducted.
Each major data element from the standard list must be modifiable by a standard list of sub-choices that further describe any abnormality. These sub-choices are not pertinent if the major element is answered “no.” A Diabetes element might have as a sub-choice list (1) insulin-dependent, (2) oral-agents, (3) diet-controlled, (4) borderline, and (5) gestational only. Lists should always include an open-ended “other” choice.
When indicated, any individual or practice can, of course, decide to collect more data than are included in the standard set . And the data collected beyond the standard can take any form.
In medical informatics terms, the objective is to be able to pool or compare data from multiple institutions to build comparisons on large numbers of patients. Pooling requires that consistently defined standard data elements are available from all participating institutions. The role of a data dictionary is to define exactly what is encompassed by each data element.
For example: what is diabetes and what is not. Does diabetes include both insulin-dependent and oral agent patients? Does diabetes include “borderline” diabetics, gestational diabetics, or diabetics who control their sugar by diet? Is hypoglycemia a form of diabetes? Anesthesia needs a data dictionary that would clearly define the limits of a yes or a no answer for each standard element. The dictionary definition also leads to the definition of sub-list choices.
Anticipation of problems, and with it, the ability to focus testing and intensive monitoring, can be derived from studies linking pre-anesthesia evaluation elements with intra- or postoperative incidents. Inclusion of pre-anesthesia evaluation data requires an understanding of the needs of medical informatics and the establishment of standard terminology and definitions for clinical use. Anesthesia clinicians must adopt structured techniques for pre-anesthesia evaluation so that data are available to support the continuing improvement of patient care.
Dr. Phillips is R&D Section Manager, Philips Medical Systems (Agilent).