Database Quality and Access Issues Relevant to Research Using Anesthesia Information Management System Data

For this special article, we reviewed the computer code, used to extract the data, and the text of all 47 studies published between January 2006 and August 2017 using anesthesia information management system (AIMS) data from Thomas Jefferson University Hospital (TJUH). Data from this institution were used in the largest number (P = .0007) of papers describing the use of AIMS published in this time frame. The AIMS was replaced in April 2017, making this finite sample finite. The objective of the current article was to identify factors that made TJUH successful in publishing anesthesia informatics studies. We examined the structured query language used for each study to examine the extent to which databases outside of the AIMS were used. We examined data quality from the perspectives of completeness, correctness, concordance, plausibility, and currency. Our results were that most could not have been completed without external database sources (36/47, 76.6%; P = .0003 compared with 50%). The operating room management system was linked to the AIMS and was used significantly more frequently (26/36, 72%) than other external sources. Access to these external data sources was provided, allowing exploration of data quality. The TJUH AIMS used high-resolution timestamps (to the nearest 3 milliseconds) and created audit tables to track changes to clinical documentation. Automatic data were recorded at 1-minute intervals and were not editable; data cleaning occurred during analysis. Few paired events with an expected order were out of sequence. Although most data elements were of high quality, there were notable exceptions, such as frequent missing values for estimated blood loss, height, and weight. Some values were duplicated with different units, and others were stored in varying locations. Our conclusions are that linking the TJUH AIMS to the operating room management system was a critical step in enabling publication of multiple studies using AIMS data. Access to this and other external databases by analysts with a high degree of anesthesia domain knowledge was necessary to be able to assess the quality of the AIMS data and ensure that the data pulled for studies were appropriate. For anesthesia departments seeking to increase their academic productivity using their AIMS as a data source, our experiences may provide helpful guidance. Accepted for publication January 29, 2018. Funding: None. The authors declare no conflicts of interest. Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's website (https://ift.tt/KegmMq). Data from this study were presented, in part, at the 2017 Annual Meeting of the American Society of Anesthesiologists, October 21–25, 2017, Boston, MA. Reprints will not be available from the authors. Address correspondence to Richard H. Epstein, MD, Department of Anesthesiology, Perioperative Medicine & Pain Management, University of Miami, Miller School of Medicine, 1400 NW 12th Ave, Suite 3075, Miami, FL 33136. Address e-mail to repstein@med.miami.edu. © 2018 International Anesthesia Research Society

from #ORL-AlexandrosSfakianakis via ola Kala on Inoreader https://ift.tt/2E5a42I