My job entails reading and summarizing numbers of clinical records. Most of them are VA and Tri Care computer generated printouts. Gleaning useful and valid information is a challenge. The printouts often lack a date of entry. They do, however, contain vast amounts of data regarding social,environmental and immunization information. It is difficult to determine which content is generated by a physician. Many designated contributors have no identity or an array of non M.D. acronims following their name. There may be pages between credible clinical notes about the patient's current condition. The most troubling EHR data is the Problem List without dates. There is begaining date to the problem and no date of its resolution. Without a date of origion, there is no way to tell who entered the problem or when. Was it a physician, or the patient stating the problem as a diagnosis to a team member not qualified to enter a diagnosis. Then too, without a date of resolution, one cannot determine if it is a chronic problem or one long since resolved. Furthermore, without a date the reader cannot find the data supporting the stated problem. Despite modern confidentiality rules, these problem lists are used to make various determinations of eligibility.
Much is written about the burden of the EHR on physician's time and the quality of the record. The EHRs are enormously expensive. They are competitive. There are many companies competing for the business. They all attempt to provide the federal requirements of meaningful use with a promise of standardized data. However, in so doing, the already standardized professional expression of patient data is lost. While the bean counters may relish the access to the data they perceive as meaningful, the information critical to patient care, continuity of care and communication between physicians is lost. There is probably no turning back, but it would seem that voice recognition with streaming physician narrative would go a long way to improve EHRs. If a problem oriented record is to be the standard, Problem Lists should be by physician input only and dated by onset and resolution. Furthermore, when multiple problems lead to a single diagnosis, there should be a simple way to edit the change or at least document the thinking. As it is, there is a lot of garbage in the system.
Comments