Medical record: systematic centralization versus secure on demand aggregation

As patients often see the data of their medical histories scattered among various medical records hosted in several health-care establishments, the purpose of our multidisciplinary study was to define a pragmatic and secure on-demand based system able to gather this information, with no risk of breaching confidentiality, and to relay it to a medical professional who asked for the information via a specific search engine.

Scattered data are often heterogeneous, which makes the task of gathering information very hard. Two methods can be compared: trying to solve the problem by standardizing and centralizing all the information about every patient in a single Medical Record system or trying to use the data "as is" and find a way to obtain the most complete and the most accurate information. Given the failure of the first approach, due to the lack of standardization or privacy and security problems, for example, we propose an alternative that relies on the current state of affairs: an on-demand system, using a specific search engine that is able to retrieve information from the different medical records of a single patient.

We describe the function of Medical Record Search Engines (MRSE), which are able to retrieve all the available information regarding a patient who has been hospitalized in different hospitals and to provide this information to health professionals upon request. MRSEs use pseudonymized patient identities and thus never have access to the patient's identity. However, though the system would be easy to implement as it by-passes many of the difficulties associated with a centralized architecture, the health professional would have to validate the information, i.e. read all of the information and create his own synthesis and possibly reject extra data, which could be a drawback. We thus propose various feasible improvements, based on the implementation of several tools in our on-demand based system.

A system that gathers all of the currently available information regarding a patient on the request of health-care professionals could be of great interest. This low-cost pragmatic alternative to centralized medical records could be developed quickly and easily. It could also be designed to include extra features and should thus be considered by health authorities.