Skip Navigation

The Value of a Rapid Anthrax Triage Tool with Minimal Need for Diagnostic Testing

Amesh A. Adalja, MD, FACP, FACEP, FIDSA, April 9, 2019

It has been 18 years since the anthrax attacks in the United States, but the threat has not vanished. The 2001 Amerithrax attack resulted in 22 people being deliberately infected and 5 dying from the infection. In response to that attack and other threats, numerous efforts to fortify the United States and its healthcare facilities against a biological attack have been pursued with the development of new vaccines, antibiotics, and other countermeasures. Additionally, programs to enhance the capacity of healthcare facilities to respond have been pursued. Part of healthcare and frontline response includes the capability to rapidly identify people who have symptoms that may or may not be attributable to a biological agent. A new triage tool to identify those who have signs and symptoms consistent with anthrax is described in an online-first article in the Annals of Internal Medicine.

Cases from 1880 to 2013

To develop the tool, Hupert and colleagues studied the clinical presentation of anthrax cases (of all types) in the English language medical literature. More than 400 patients were included; they were compared with 653 control patients who were seen for “anthrax-related concerns” at 2 emergency departments in Virginia and New York, considered the epicenters during the Amerithrax attack. Multivariable regression was performed to identify specific constituents of the checklist.

The 407 anthrax patients included a majority with cutaneous anthrax (59% with systemic manifestations) and 19.1% with inhalational anthrax.

Checklist with 10 Elements

Multivariable regression analysis identified altered mental status, tachypnea, fever, dyspnea, skin lesions, abnormal lung or heart sounds, increased girth (an indication of ascites), absence of cold symptoms, and hypothermia as being independent predictors of anthrax. Case patients were also noted to be more febrile, more tachypneic, and more tachycardic than controls.

The final checklist included heart rate, respiratory rate, temperature, mental status, diaphoresis, dyspnea, severe headache, characteristic skin lesions, abnormal lung sounds, and increased girth. Diaphoresis and severe headache were pulled from univariate analysis.

The checklist was developed to have 2 aspects: paramedical and medical. The paramedical portion of the triage checklist, which could be assessed by individuals with lower levels of training, would trigger antimicrobial treatment if any of the following were present: heart rate ≥110 beats per minute, respiratory rate >20 breaths/minute, altered mental status, or severe headache. Characterization of auscultation findings, skin lesion identification, and presence of ascites require a higher level of expertise and therefore are part of the medical portion of the checklist. Less significant scores might trigger medical evaluation or postexposure prophylaxis. The checklist correctly classified 95% of case patients, with the remaining 5% being identified as needing postexposure prophylaxis, with instructions to return for worsening symptoms.

Simple but Valuable

This triage tool is a significant development because it minimizes the need for what would be, in a biological attack environment, cumbersome diagnostic testing. The chief strength of this checklist approach is that it is simple, requires little training, and provides clear guidance as to what path a patient is on (treatment, medical evaluation, postexposure prophylaxis). The fact that the tool is derived from actual case patients (having all forms of anthrax) matched to anthrax-concerned controls adds to its validity.

As the tool gains traction, it might be necessary to see how many cases of non-anthrax-related sepsis it includes, as some of the tool’s elements bear similarity to qSOFA scoring used for sepsis identification. However, sepsis patients need urgent medical attention and antimicrobial therapy as well, so identifying them with the tool may not be a major drawback. The tool would be expected to be deployed only with specific triggers.


Hupert N, Person M, Hanfling D, Traxler RM, Bower WA, Hendricks K. Development and performance of a checklist for initial triage after an anthrax mass exposure event. Ann Intern Med 2019; Accessed April 8, 2019.