Affiliation:
1. Assistant Professor.
2. Professor, Departments of Anesthesiology and Obstetrics, Gynecology, and Reproductive Sciences, Mount Sinai School of Medicine.
3. Programmer Analyst.
4. Senior Faculty.
5. Professor and Chair, Department of Anesthesiology.
Abstract
Background
Anesthesia information management systems allow automatic recording of physiologic and anesthetic data. The authors investigated the prevalence of such data modification in an academic medical center.
Methods
The authors queried their anesthesia information management system database of anesthetics performed in 2006 and tabulated the counts of data points for automatically recorded physiologic and anesthetic parameters as well as the subset of those data that were manually invalidated by clinicians (both with and without alternate values manually appended). Patient, practitioner, data source, and timing characteristics of recorded values were also extracted to determine their associations with editing of various parameters in the anesthesia information management system record.
Results
A total of 29,491 cases were analyzed, 19% of which had one or more data points manually invalidated. Among 58 attending anesthesiologists, each invalidated data in a median of 7% of their cases when working as a sole practitioner. A minority of invalidated values were manually appended with alternate values. Pulse rate, blood pressure, and pulse oximetry were the most commonly invalidated parameters. Data invalidation usually resulted in a decrease in parameter variance. Factors independently associated with invalidation included extreme physiologic values, American Society of Anesthesiologists physical status classification, emergency status, timing (phase of the procedure/anesthetic), presence of an intraarterial catheter, resident or certified registered nurse anesthetist involvement, and procedure duration.
Conclusions
Editing of physiologic data automatically recorded in an anesthesia information management system is a common practice and results in decreased variability of intraoperative data. Further investigation may clarify the reasons for and consequences of this behavior.
Publisher
Ovid Technologies (Wolters Kluwer Health)
Subject
Anesthesiology and Pain Medicine
Cited by
16 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献