Natural Language Processing and Quality Assessment in Primary Care

Diabetes

foot exam, health services, practice improvement, quality, preventive services, Natural Language Processing

The investigators propose to use a Natural Language Processing System (NLP) to provide an initial baseline report for primary care patients at risk for diabetes and cardiovascular complications that will include: a) evidence of foot exam documentation in the previous year; b) use of aspirin for cardiovascular risk reduction; and c) tobacco use. As part of a randomized trial, we plan to use a previously validated mailed survey (NCQA Provider Recognition Program) that requests information on the last foot exam, use of aspirin and tobacco. Patients who have been identified by NLP as not having had a foot exam will be randomized into treatment and control arms. Both arms will receive an informational letter; with a second mailing to nonresponders after one month, describing the key strategies for effective patient-physician communication during the clinical encounter. The treatment arm will also receive an informational letter and patient education brochure containing key messages about the importance of regular foot examinations. NLP will be repeated after 6 months to compare the impact of the patient education materials.

No related papers found yet.

United States