Each year, the medical evidence piles inexorably higher. Even though it’s literally impossible to keep up, we still try. Without further ado, a short list of new developments from 2019:

What to Do When Benzodiazepines Fail

This year saw the publication of several studies involving second-line treatment for status epilepticus when patients’ seizures are refractory to benzodiazepines. There has been a general shift toward using levetiracetam (Keppra), likely due to ease of administration and perceived advantages implicit to its newness. Three pediatric studies tested levetiracetam against other second-line agents.^1–3 Two of these studies were head-to-head comparisons against phenytoin, and one added a third arm featuring valproic acid. Across all three studies, despite minor variations in secondary outcomes, no clear “winner” was found. An individualized choice of any of these agents may be considered reasonable while we await further developments in antiepileptic therapy.

The State of Pulmonary Embolism (PE) Exclusion

Thankfully, we are continuing to make progress toward reducing the use of advanced imaging in the evaluation of PE. Two studies published in the past year illustrate potential strategies to address imaging overuse.^4,5 The first looks at the use of the YEARS protocol for the evaluation of PE in pregnant women, using the combination of high-risk features and two different D-dimer thresholds to increase the number of women we can safely conclude do not require imaging. While this prudent application of YEARS was shown to improve imaging stewardship, it also illuminated the regrettable over-triage of pregnant women to evaluation for PE and an underlying baseline culture of pervasive advanced imaging. Further useful work in this field may incorporate trimester-adjusted D-dimer in addition to further decision support.

A second study parallels the YEARS concept except it uses the Wells Score as the foundation, dividing the cohort into low-, intermediate-, and high-pretest likelihood for PE. The D-dimer Testing Tailored to Clinical Pretest Probability in Suspected Pulmonary Embolism (PEGeD) study doubled the D-dimer imaging threshold cutoff for patients with a low-pretest likelihood, and no cases of missed PE were observed. While this is a successful demonstration of their strategy, considering clinical equipoise for PE allows a miss rate of around 1 percent to balance harms and benefits, even more aggressive strategies are likely reasonable. At the least, this study still represents another important step further establishing pretest-adjusted D-dimers as appropriate.

Finally, the most concise update is from the realm of syncope. Several years ago, a systematic evaluation of hospitalized patients with syncope found a prevalence of PE of 16 percent.⁶ Now, another prospective study finds the prevalence of PE at presentation (which is mainly what we care about when making decisions about who to work up) to the emergency department to be much lower: only 1.4 percent in these data.⁷ An evaluation for PE is necessary only as otherwise clinically indicated in the context of syncope.