Background and aims

Sore throat is very common in children. It can be caused by viruses or bacteria. The bacterium most frequently identified during sore throat in children is group A streptococcus ('strep throat'). Amongst children with sore throat, antibiotic treatment is only useful in those with strep throat.

Simple, rapid tests for the diagnosis of strep throat have been available since the 1980s. Physicians can do a rapid test at the point of care by swabbing the throat. Based on the result of the rapid test, they can then decide if antibiotics are needed.

We reviewed the evidence about the performance of rapid tests for correctly detecting strep throat in children seen in Outpatient departments with a main complaint of sore throat.

Study characteristics

We searched for studies published in any language from January 1980 to July 2015. We found 98 unique studies, for a total of 116 test evaluations, involving 101,121 children. The number of participants ranged from 42 to 11,644 across test evaluations. The proportion of children with strep throat ranged from 9.5% to 66.6% across test evaluations.

Quality of the evidence

Important study design features were frequently not reported. The overall methodological quality of included studies was poor. For most studies, we had concerns about the ways in which participants were selected.

Key results

On average, rapid tests for strep throat had a sensitivity (ability to correctly detect people with the disease) of 86% and a specificity (ability to correctly identify people who do not have the disease) of 95%. There was substantial variability in rapid test performance across studies, which was not explained by study characteristics, including methodological quality. The two types of rapid tests under evaluation seemed to have comparable sensitivity (85.4% versus 86.2% for enzyme immunoassays and optical immunoassays, respectively). Based on these results, we would expect that amongst 100 children with strep throat, 86 would be correctly detected with the rapid test while 14 would be missed and not receive antibiotic treatment. Of 100 children with non-streptococcal sore throat, 95 would be correctly classified as such with the rapid test while 5 would be misdiagnosed as having strep throat and receive unnecessary antibiotics.