On-Line Medical Record

This graph shows the steady growth of OLMR use between June 1992 and June 1995. Each of the five categories represents an OLMR function that returns patient specific data. The "Historical" and "Document" functions account for the majority of OLMR queries. Clinicians use the historical view to quickly review a patient's complete database holdings. These lists are conveniently sorted in reverse chronological order so that recent additions are always at the top. Administrative staff are often interested in specific information related to admission, discharge, or billing. They tend to use the document view to filter out extraneous data. The historical view includes lab data and this may explain the relative underuse of the lab specific functions. There are also two alternative sources for these data, the Laboratory and Hospital Information Systems.

Current Database Holdings

The OLMR currently holds over 30 million lab results and 1.5 million documents. The database is cumulative; older records have not been purged due to lack of storage capacity. Data are collected automatically from most ancillary systems by a microcomputer interface called "Input Manager." A simple but effective user interface allows access from hard-wired 3270 terminals or networked desktop computers. Access from MS-DOS, MS-Windows, Apple Macintosh, and UNIX systems is possible using terminal emulation. An example OLMR session is shown on the next page.

Example of OLMR Use

While the number of transcribed documents has increased steadily over the past three years, actual filing of these documents has not kept pace. Less than one third of clinic notes were available in the paper chart by the time of each subsequent visit. By virtue of a direct upload path, 100% of transcribed documents were available in the OLMR. In June of 1995 the OLMR was the only source of prior clinic notes for 47% of return visits. The number of handwritten notes has decreased significantly over the past year as OLMR availability has increased.

This graph shows availability of clinic notes as a percentage of total return visits. The first bar in each pair shows the actual availability encountered in the study. The second bar shows the same data as if the OLMR did not exist. Several conclusions can be drawn from this comparison:

• In spite of increased transcription, access to clinic notes in paper form is largely inadequate. The OLMR is essential to the orderly and timely flow of information in the adult oncology clinic.
• As the availability of transcribed documents improves, clinicians may be less inclined to hand write their clinic notes.
• The number of return visits for which there is no data available has decreased from approximately 10% to almost zero over the course of the study.
• There is a small, recalcitrant number of visits for which the clinical data available are incomplete. The cause of this problem was not obvious from our analysis and suggests an area for future study.