Medical algorithms in breast pathology — An implementation and evaluation

Medical Algorithms (MAs) are tools (any computation, formula, survey or look-up table) that aid physicians in clinical decision making. A central Portuguese hospital has a Virtual Electronic Patient Record that includes a Breast Pathology module (VCBreastCare) being used for the last 3 years. Aim. We aim to (1) select and implement MAs in VCBreastCare and (2) evaluate the use of these by its' users. Methods. First, a list of algorithms was retrieved through a site search in The Medical Algorithm Project (Medal), and experts selected 13 medical algorithms. An extra algorithm considered important that was not in Medal was included. A questionnaire to evaluate the importance of each of the 14 MAs was applied to the VCBreastCare users and the 3 most important MAs were implemented using Web technologies. The evaluation of their use was performed by another questionnaire and by analyzing the MAs usage logs during the first 3 months. Results. MAs for “Family History”, “Breast Biopsy” and “TNM Classification” were implemented and made available in September 2011. Although the physicians initially considered MAs important, they did not use them significantly (<; 1% of patients). Discussion. The low usage rate was disappointing when compared with their initial motivation. Four explanations/measures can now be proposed: (1) due to the high quality of the health professionals involved MAs are simply not useful, (2) the selected MAs were not the most appropriate, and so new MAs will be implemented, (3) automate MAs, thus the result is automatically calculated and alerts are displayed indicating consults or exams that patient should do, or (4) users will be forced to use the MAs by incorporating them in the patient record.