UNC Hospitalist and Professor of Medicine Dr. John Stephens has authored a new perspective in the Journal of Hospital Medicine series: Things We Do for No Reason. The series seeks to highlight practices in clinical medicine that are common but either have poor evidence to support them or actual evidence against the practice.

Delirium is a common diagnosis in hospitalized patients, associated with longer stays and higher mortality, estimated to cost over $164 billion per year in the US.

Stephens, with representatives from Atrium Health and Duke, focused on neuroimaging for hospitalized patients with delirium because of their own shared experiences seeing head CTs and brain MRIs frequently ordered for delirious patients with a low rate of important clinical findings.

John R. Stephens, MD

CT imaging exposes patients to significant radiation, and up to 2% of malignancies in the US may be attributed to prior tomography exposure. Additionally, sedating medications are often administered to agitated patients prior to imaging, which can worsen delirium. Ordering neuroimaging for all patients with acute delirium, therefore, exposes the large majority to unnecessary costs and potential harms.

“Our beliefs were reinforced by a review of the literature,” said Stephens. “We found that in the absence of specific risk factors, such as a recent fall, being on anti-coagulation, or the identification of focal neurologic findings on exam, the yield of head CT for important intracranial pathology was only about 1%.”

The perspective identifies what should instead be done to evaluate delirium in hospitalized patients, and suggests that a validated delirium assessment tool be used to evaluate hospitalized patients who develop altered mental status. The recommendation also says that patients should undergo a thorough history aimed at identifying common risk factors. It recommends neuroimaging be performed only if there is a history of a fall or head trauma in the preceding two weeks, if there are any focal abnormalities, or if the patient is receiving systemic anticoagulation. Finally, the perspective suggests neuroimaging may be a reasonable consideration for patients with an atypical course of delirium.

“Recognizing that the yield of head CT pathology is only about 1% provides us with an opportunity.  We’d like to use this data to spur further study at our institutions, ideally to develop a validated risk prediction tool that can help to further identify low-risk patients who can be spared the costs and harms of unnecessary imaging.”

The article was published March 20, 2019, in the Journal of Hospital Medicine.  Other authors included Stephen Chow, DO, Andrew McWilliams, MD, and Daniel M Kaplan, MD.  Find the article here.

Inspired by the American Board of Internal Medicine’s Choosing Wisely® campaign, the “Things We Do for No Reason” (TWDFNR) series reviews practices that have become common parts of hospital care but may provide little value to our patients. Practices reviewed in the TWDFNR series do not represent “black and white” conclusions or clinical practice standards but are meant as a starting place for research and active discussions among hospitalists and patients.