In Reply We thank Dr McDonald for his observations on our studies and agree that primary care practitioners (PCPs) deserve more signal and less noise through their electronic health record (EHR)-related alerts. Primary care practitioners must also agree on the signal itself—what types of information they want or need to receive. We have found that physicians often disagree about the importance or urgency of alert content.1 Dr McDonald’s example (ie, transmission of abnormal vs all test results) offers one possible point of consensus. However, in our experience of developing policy on communicating test results within the Department of Veterans Affairs (VA),2 we found this rather challenging to operationalize. For instance, an alert should be generated for abnormal findings that are likely unexpected (eg, when the history given to the radiologist is “rule out cancer” or “cough”), but PCPs need not be alerted to abnormalities that are clearly known to them (eg, unchanged cancer metastases on a computed tomography request for “follow-up of metastatic lung cancer”). Conversely, certain “normal” laboratory results are essentially not normal in the given patient’s context (eg, normal international normalized ratio in a patient receiving warfarin). Thus, “normal” and “abnormal” as conventionally defined do not necessarily map onto a simple protocol for efficient use of alerts. Physicians need to come to an agreement on what defines an “alert worthy” result and how to most effectively handle the inevitable exceptions to the rule. Moreover, PCPs must be given flexibility to customize receipt of certain types of alert notifications that are not considered “alert worthy” by others.
Singh H, Sittig DF. Toward Electronic Medical Record Alerts That Consume Less Physician Time—Reply. JAMA Intern Med. 2013;173(18):1756. doi:10.1001/jamainternmed.2013.9317