Automated data mining: an innovative and efficient web-based approach to maintaining resident case logs.

Area of Science:

• Neurology
• Medical Education
• Data Science

Background:

• The Accreditation Council for Graduate Medical Education (ACGME) has considered resident case logs for neurology residency programs.
• Manual case logging is a current practice but has limitations.

Purpose of the Study:

• To evaluate the effectiveness of a data-mining program for generating resident case logs.
• To compare data mining with manual data entry systems for resident logging.
• To explore additional applications of data mining in resident education.

Main Methods:

• Patient notes were extracted from the Hospital Information System.
• An unstructured data-mining program was used to analyze resident dictations.
• Extracted data included history, examination, and ICD codes, compared against manual logs over 30 days.

Main Results:

• The automated method successfully extracted dictations from all 19 residents, while only 4 used manual logging.
• Manual logs identified broad categories (e.g., stroke, epilepsy, pain), whereas data mining provided specific diagnoses (e.g., epilepsy, cerebral atherosclerosis, headache) and chief complaints (e.g., headache, seizures, low back pain).
• Data mining extracted detailed patient information like tobacco, alcohol, and drug use, which manual logs did not capture.

Conclusions:

• Manual case logs are inefficient, offer limited data, and may be unpopular with residents.
• Data mining is a time-efficient tool for assessing resident experience and ACGME core competencies.
• Further research on data mining in larger residency programs is recommended.