Performance of an algorithm for assessing smallpox risk among patients with rashes that may be confused with smallpox.

After the 2001 anthrax bioterror attacks, the Centers for Disease Control and Prevention developed an algorithm to evaluate patients rapidly for suspected smallpox. A prospective, multicenter study examined the performance of this algorithm in assessing patients with an acute, generalized vesicular or pustular rash (AGVPR) admitted to emergency departments and inpatient units of 12 acute-care hospitals in 6 states. Of 26,747 patients (3.5% of all admissions) with rashlike conditions screened, 89 (1.2 patients per 10,000 admissions) had an AGVPR. Physicians or study staff classified none of 73 enrolled patients as being at high risk for having smallpox; 72 (99%) were classified as being at low risk, and 1 was classified as being at moderate risk. The discharge diagnosis for 55 (75%) of these 73 participants was varicella illness. Use of the algorithm did not result in misclassification of AGVPR as high risk for smallpox. The algorithm is a highly specific tool for clinical evaluation of suspected smallpox disease.