Do We Need Specific Certification to Use Anesthesia Information Management Systems?

Abstract

To the Editor We appreciated the recent publication by Simpao and Rehman1 discussing Anesthesia Information Management Systems (AIMS). As the authors suggest, machine learning could be applied to AIMS data to improve extraction, assign meaning or clinical relevance, and contribute to clinical decision support systems.1 However, these “augmented” technologies also make the systems more complex and sophisticated. Consequently, obtaining recognized certifications in information technologies appears crucial to ensure optimal use that appreciates both its potential and limits.2 Moreover, it is possible that obtaining certifications would convince administrators to allow greater autonomy in data management, even those not strictly involved in the anesthetic department.3 However, as emphasized by Hofer et al,3 the specificity of anesthetic data could result in suboptimal clinical informatics training if it is not performed in an anesthetic department. We believe that in an anesthesiology residency program, specific modules in “informatic anesthesia” should be implemented to allow the new generations to interface immediately with those systems. Another fundamental aspect for their correct use is the quality of data, as the authors also point out.1 Among possible implementations that could help is the minimization of manual introduction of data. Phelps et al4 demonstrated that distribution of the start and end anesthesia times was more uniform with AIMS than paper records. In addition, different interfaces, which further reduce manual input, such as scanning barcodes or button clicks, could improve data faithfulness. Last, we support the great potential of AIMS application for research purposes. Unfortunately, an important problem has been detected and is represented by the presence of artifacts. Subtypes of artificial intelligence could recognize them, thanks to specific algorithms. We also can imagine a study identified by codes specific to its research protocol that, once inserted into the system and applied to the specific patient, would automatically and prospectively collect the required data and fill in the case report forms. This would guarantee simplified and extremely precise clinical data collections and limit human bias and variability. In conclusion, we see great potential in AIMS. These technologies could allow accurate and incomparable quality data collections. In addition, powerful algorithms capable of answering simple questions could provide valid support in clinical decision making. The increasing complexity of the systems, however, makes certification essential in clinical informatics. Elena Bignami, MDValentina Bellini, MDAnesthesiology, Critical Care and Pain Medicine DivisionDepartment of Medicine and SurgeryUniversity of ParmaParma, Italy[email protected]