DITTO - a tool for identification of patient cohorts from the text of physician notes in the electronic medical record.

A number of important applications in medicine and biomedical research, including quality of care surveillance and identification of prospective study subjects, require identification of large cohorts of patients with a specific diagnosis. Currently used methods are either labor-intensive or imprecise. We have therefore designed DITTO - a tool for identification of patients with a documented specific diagnosis through analysis of the text of physician notes in the electronic medical record. Evaluation of DITTO on the example of diabetes mellitus, hypertension and overweight has shown it to be rapid and highly accurate. DITTO processed 170,000 notes/hr with sensitivity ranging from 74 to 96%, and specificity from 86 to 100%. Its accuracy substantially exceeded the performance of currently used techniques for each of the three diseases. DITTO can be adapted for use in another healthcare facility or to detect a different diagnosis. DITTO is an important advancement in the field, and we plan to continue to work to enhance its functionality and performance.