CEO SUMMARY: Pathologists at the University of Minnesota Medical School are working to create an electronic data base that covers the more than 100 years of autopsy cases that have been archived and stored. Their goal is to use this information to improve teaching and to further research into the evolving nature of many diseases.
ACADEMIC PATHOLOGISTS at the University of Minnesota Medical School in Minneapolis are creating an electronic data base that will eventually include more than 100 years of autopsy reports, slides, and even tissues.
The pathology department expects this data base will be used for teaching and in consultative conferences. More intriguing, pathologists at the University of Minnesota (U.M.) believe the autopsy archives may have great value in a variety of research applications, including genomics.
Earliest Cases Date To 1899
“These archives consist of autopsy reports dating from 1899, glass slides dating from 1910, and paraffin blocks of tissue dating from 1940,” stated James Coad, M.D., Professor of Pathology at U.M. “Essentially, we have a reasonably continuous record of autopsies extending back 102 years.
“More importantly, because U.M. is an academic center, these tissues represent cases referred here because they were unusual, examples of a rare disease state, or undiagnosable to the referring physicians,” he explained. “Some of these tumors are rare, with less than one or two cases referred here in a typical year.
“Starting six months ago, we embarked on a project to scan and enter the autopsy reports into an electronic data base,” said Dr. Coad. “This permits us to do key word text searches of both the diagnosis and the entire report.
“Because medical terminology has changed over the decades, the key word search makes it easier and more accurate to locate the relevant cases for teaching purposes and research,” he added.
“Until we finish scanning all the autopsy reports, it’s tough to estimate the total number of cases in the archive,” stated Ike Castrodale, the computer programmer supervising the project. “Since the turn of the last century, anywhere from 100 to 500 autopsy cases per year were archived. The total number may range from 25,000 to 35,000 individual cases.
“To this point, we’ve entered approximately 10 years worth of cases into the data base,” commented Dr. Coad. “It’s only been recently that we’ve begun to teach the pathologists how to work with the data base. Use of the data base is increasing and our pathologists are learning how to identify relevant cases with more speed and accuracy.”
The paraffin blocks of tissue represent one of the most intriguing resources in the archives. According to Dr. Coad, the University of Minnesota Medical school began to retain and archive paraffin blocks of tissue starting in 1940.
“Depending on the state of preservation, these paraffin blocks will allow us to do a limited analysis of genotypes and phenotypes,” noted Dr. Coad. “These tissues offer us the potential to consider whether different lifestyles, diets, and environments of earlier decades might play a role in how diseases evolve over time.
“For example, can these tissues help us to learn whether the prostate cancer of 2001 is different from prostate cancer in 1920?” he asked.
Evolution Of Disease
“Moreover, these specimens would represent the effects of different modes of treatment,” continued Dr. Coad. “For teaching purposes, we can show medical students advanced-stage tumors that now are seldom seen because of early detection and effective treatment. For research purposes, these tissues may help us identify how different treatment methods have shaped or altered diseases as we see them today.”
Pathologists at the University of Minnesota seem to be among the pioneers at creating an electronic data base of autopsy archives that extend back many decades. Dr. Coad notes that a handful of academic centers have put some data on-line, indexed by SNOMED codes. But more complete information on the indexed cases is generally not accessible from these sites.
“We are in the early stages of cataloging and mining data in raw form,” he offered. “As we get more years’ worth of cases into the data base, it will become increasingly useful and valuable. This data base will make it easier to utilize what are now secondary resources—the glass slides and tissues.
“Like many similar projects, as we learn more about how to use the data base, pathologists here are discovering new applications for this data,” observed Dr. Coad. “It shows how a well-constructed data base contributes to better use of the component information. Instead of spending hours looking through microfiche or bound volumes of case reports, we can now sit at a computer and quickly identify the information we need.”
THE DARK REPORT believes that the 100 years of clinical data contained in the U.M. autopsy archives represent immense value to the academic center pathology group. To its credit, it is willing to invest the resources necessary to digitize that data, making it accessible for study.
Even though this academic group is focused on its teaching mission, by digitizing its archives, it is taking the important first step to converting this raw data into information which has value to both the University and the healthcare community.
The interesting follow-up to this story will be how pathologists at the U.M. develop applications for these cases in fields such as genetics and molecular pathology.