Some doctors have a knowledge problem in proportion to collect valid data in unison to clinical research protocols wich is detached from the health record. The international terminology model SNOMED CT (SCT) is suitable to connect to the health record to data extraction. Though knowledge is missing on methods to implement the teoretical potentials about SCT in a research protocol can be used. On this basis the challenges related to development of a design with the usage of SCT to connect two types of document health record and the detached protocol regarding creating a valid data extraction in clinical trials are studied.
This is studied by first making an analysis of a existing protocol whereafter a qualification of design by connecting it to the knowlegde about SCT related to the data extraction. Finally a verification of the template design takes place by the protocol owner through a combination of think aloud test and interview. The knowledge collection is amongst other done through literaturestudies.
A prototype on a SCT template is designed and the were some challenges in order to connect the classification systems wit SCT. The coverage of the template is 98,8 % but it was only possible to map 69 out of 81 with precoordinated concepts.
The challenge lies in keeping backward compatibility if a protocol is designed entirely on SCT as a replacement for the classification systems.
This study contributes with new knowledge to create validity in the data which is extracted to clinical trials by using SCT. Results validation of the templates is at first not represenative because they are collected retrospective during the think aloud test but on the other hand it does not have any vallue in order to understanding the concepts which are chosen to be meaningful i relation to the context they are used in.