CLINDOC - a Software Tool for Detection of Clinical Inertia in Management of Hypertension through Analysis of Free Text Physician Notes
Alexander Turchin Partners HealthCare System, Inc. Merri L Pendergrass Brigham and Women's Hospital Jonathan S Einbinder Partners HealthCare System, Inc.

Introduction Delay in making treatment changes (clinical inertia) is a common problem. Currently clinical inertia is detected by chart review, which is impractical on a large scale. We developed CLINDOC - a software tool that analyzes free text physician EMR notes to identify clinical inertia in management of hypertension. Methods CLINDOC uses pattern matching to identify elevated blood pressure values in the note and detect documentation of a treatment change (starting a new or changing the dose of an existing anti-hypertensive medication). To evaluate CLINDOC performance, 300 randomly selected physician notes were analyzed by CLINDOC and by manual review. Concordance between CLINDOC and manual review results for a) blood pressure values recorded in the note, b) documentation of an appropriate treatment change if blood pressure was elevated was determined. . Results CLINDOC analyzed 226,000 notes/hour. Agreement between CLINDOC and manual review for blood pressure was 97.3% (kappa 0.963, sensitivity 95.3%, specificity 98.0%). Agreement between CLINDOC and manual review for medication adjustment was 89% (kappa 0.783, sensitivity 73.5%, specificity 97.0%). Conclusions CLINDOC rapidly and reliably detects clinical inertia in the management of hypertension. Its speed and accuracy are sufficient for enterprise-wide monitoring and research of the phenomenon of clinical inertia.