Abstracts

Rationale: Computerized neurodiagnostic equipment for long-term monitoring and EEG have evolved from stand-alone systems with local storage and media archiving, toward networked systems which allow remote access and monitoring. In the past, these systems typically used private networks especially for video-EEG applications that hospital backbones were generally unable to carry. Modern healthcare facilities now have high-speed backbones with bandwidths in the gigabyte/sec range and this facilitates centralization of data management. Here we describe our experience and the advantages of moving to centralized data center storage for our neurodiagnostic data. Methods: Our neurodiagnostic equipment consists of 12 EEG/LTM recorders with digital EEG and video capabilities installed in 2 dedicated monitoring units and the EEG laboratory. Additionally, we utilize 2 networked portable recording units and 17 EEG review stations located throughout the facility. In 2008, we transitioned from 3 local, self-maintained file servers and local private networks to a centralized file server maintained and supported by our IS department. Although EEG data and digital video is recorded directly to the file server, we maintain the ability to write locally on each recorder in the event of a network or server failure. This is part of an overall program to centralize all medical records in a standardized, secure, and cost-efficient way at our medical center. Results: The major pre-requisite to this transition was the installation of a high-speed backbone capable of reaching locations where monitoring and data review are performed. The transition went smoothly over the course of several days during which the recorders were redirected to write through the hospital network to the centralized file server as opposed to the local file server. No degradation of performance in writing, accessing or archiving was encountered. In addition, remote access from other areas of the medical center was improved, in particular when reviewing digital video. The new server offers relatively unlimited storage capacity which allows archived data to remain on-line indefinitely. We no longer perform media backups (e.g. CD, DVD) of our data, and instead rely on automated backup and data retrieval procedures for data security. Accidental data loss during the editing process is still possible and has occurred, but in those few instances, we have been able to restore data successfully with the help of our IS colleagues. The cost of centralized data storage is estimated to be approximately $5/gigabyte, which is very modest, considering the advantages of the approach. Conclusions: This approach is secure and cost-effective, and frees neurodiagnostic staff from technical responsibilities which are often beyond their training. We are considering doing the same for other neurophysiological (EP, IOM), imaging and patient information data that we work with. Distributed data access, including off-site access by clinicians raises security issues that will need to be addressed, but this is a generic IT problem for which solutions already exist.