Table 2

Summary estimates of sensitivity, specificity, positive likelihood ratio (LR) and negative LR calculated for signs and symptoms using a bivariate random effects model

Sign or symptom

No. of studies

No. of patients

Sensitivity (95% CI)

Specificity (95% CI)

+LR (95% CI)

-LR (95% CI)


Absence of cough

19

4,653

0.74 (0.68 to 0.79)

0.49 (0.40 to 0.58)

1.46 (1.28 to 1.66)

0.53 (0.46 to 0.61)


Fevera

21

4,635

0.50 (0.39 to 0.62)

0.70 (0.58 to 0.79)

1.65 (1.40 to 1.95)

0.71 (0.64 to 0.80)


Anterior cervical adenopathyb

9

2,101

0.65 (0.55 to 0.74)

0.55 (0.45 to 0.64)

1.45 (1.25 to 1.67)

0.63 (0.52 to 0.76)


Tender anterior cervical adenopathyb

16

4,144

0.67 (0.52 to 0.79)

0.59 (0.49 to 0.69)

1.65 (1.41 to 1.92)

0.56 (0.41 to 0.76)


Any exudatesc

21

4,839

0.57 (0.44 to 0.70)

0.74 (0.63 to 0.82)

2.20 (1.76 to 2.74)

0.58 (0.47 to 0.72)


aThe most widely used cut-off point to indicate fever was 38.0°C. Studies also used 37.5°C [40], 37.8°C [37,44,46], 38.3°C [13], 38.5°C [31,45].

bStudies reported on tender lymph nodes, adenopathy or tender adenopathy. All adenopathy results were categorised into tender anterior cervical adenopathy or anterior cervical adenopathy.

cIncludes results for 'tonsillar exudate', 'pharyngeal exudate' and 'exudate'. If both tonsillar and pharyngeal exudate were reported only the results for tonsillar exudate were added to avoid double counting.

Aalbers et al. BMC Medicine 2011 9:67   doi:10.1186/1741-7015-9-67

Open Data