Using natural language processing technology in electronic health record systems has helped identify patients with multiple sclerosis and collect information on disease traits, according to a study by researchers at Vanderbilt University Medical Center, Health Data Management reports.
Details of the Study
The study — published in the Journal of the American Medical Informatics Association — identified 5,789 patients with MS by using information from their EHRs to create an algorithm. The algorithm included data from:
- ICD-9 codes;
- Medications; and
- Text keywords.
Researchers also collected data on the clinical course of disease progression.
According to the study’s authors, “This is one of the first studies to focus on specific traits of a disease by text mining of the [EHR].”
The study found that for all clinical traits examined:
- Precision was 87%; and
- Specificity was greater than 80% (Goedert, Health Data Management, 12/7).
Reaction
The researchers wrote , “This dataset provides a rich resource for better understanding MS and also shows that extraction of detailed disease states and markers of prognosis in patients with chronic disease is possible and may yield a powerful tool in chronic disease research.”
They added, “This information is extractable from clinic notes by simple algorithms, with high specificity, precision, and recall”