Skip Navigation

header

Biosurveillance and Group A Streptococcal Pharyngitis

By Amesh A. Adalja, MD, FACP, October 21, 2011

The likelihood that a clinical diagnosis of a communicable disease will be accurate is influenced by the prevalence of the disease in the community: The higher the prevalence, the more likely the clinical diagnosis will be accurate. However, clinicians often do not have access to real-time prevalence data. Fine et al.1 recently published a retrospective study to determine whether access to real-time prevalence data (ie, biosurveillance) for the very common condition of group A streptococcal (GAS) pharyngitis could potentially improve clinical care if incorporated into treatment algorithms.

Centor Score Coupled with Biosurveillance-Responsive Score

In this retrospective analysis, the widely used clinical scoring rule known as the Centor score was coupled to a novel parameter computed using local prevalence data in patients 15 years and older who had attended an urgent care clinic and received testing for GAS pharyngitis. The study population was composed of more than 130,000 patient encounters spanning 6 states. Table I details the elements of the Centor score. Table II summarizes the Centor score’s recommendations based on score.

Table I: Elements of the Centor Score (1 point for each criterion met)

History of fever

Tonsillar exudates

Tender anterior cervical adenopathy

Absence of cough

Table II: Centor Score and Treatment/testing

Score

Action

0

No testing or treatment for GAS pharyngitis

1

No testing or treatment for GAS pharyngitis

2

Test for GAS pharyngitis, treat if positive

3

Treat for GAS pharyngitis

4

Treat for GAS pharyngitis

The authors created a biosurveillance variable that reflected disease incidence. The recent local proportion positive (RLPP) parameter was defined as the proportion of patients testing positive out of all the patients tested for GAS pharyngitis in the local area in the previous 2 weeks. The rate of positive GAS tests in those with specific Centor scores was calculated for different prevalence (RLPP) levels.

RLPP Can Modify the Centor Score

Not surprisingly, analysis of the dataset revealed that when the RLPP in a given area increased—reflecting a higher prevalence of GAS pharyngitis—the proportion of positive GAS tests also increased for those with Centor scores of 1 to 4.

Fine et al. modified the scores by adding 1 point to those with Centor scores of 1 if the RLPP in a given area was above 0.3—a change that could alter management. When this modification was made in the retrospective dataset, 139 additional patients were correctly classified as positive for GAS, while only 41 were incorrectly identified as positive. Extrapolating these results to national estimates, more than 62,000 additional cases of GAS could potentially be uncovered by increased testing. Similarly, subtracting a point from Centor scores of 3 when the RLPP was below 0.20 (ie, testing before treating them) could potentially spare 5 million antibiotic doses when extrapolated nationally, because this cohort of patients would receive tests before treatment decisions.

The Potential for Real-Time Biosurveillance to Affect Real-Time Patient Care

The results of this study provide encouraging evidence that implementing real-time biosurveillance could potentially have an impact on how a clinician treats an individual patient. The results of this retrospective data analysis support more judicious antimicrobial use and pathogen testing through simple data gathering and dissemination. While this study focused on GAS pharyngitis, the concept could be applied to other conditions, including influenza, RSV, rotavirus, urinary tract pathogen antimicrobial susceptibility, and sexually transmitted infections.

Clinicians often do not have timely access to the percent positivity of microbiological testing and are unable to benefit from what clinicians the day prior (or even the hour prior) may have discovered about the prevalence of various conditions in their region. The growing use of health information technology and the ability to quickly access data electronically can make it feasible for clinicians to have access to this information.

References

  1. Fine AM, Nizet V, Mandl KD. Improved diagnostic accuracy of group Astreptococcal pharyngitis with use of real-time biosurveillance. Ann Int Med 2011;155:345-352.