Table 4

Analysis of timing and site of swabs.

Timing (36 v 32 weeks)

Site (Either Lower Vaginal or Perianal vs Lower Vaginal Swab alone)

Difference (36 – 32)

95% Confidence Interval

p-value

Difference (Either vs LVS)

95% Confidence Interval

p-value

Sensitivity

8.8%

(1.2%, 16.3%)

0.023*

8.2%

(4.6%, 11.7%)

<0.001†

Specificity

0.6%

(-1.2%, 2.5%)

0.493

-1.8%

(-2.6%, -0.9%)

<0.001§/P >

Negative Predictive Value

2.3%

(0.3%, 4.3%)

0.026*

2.0%

(1.0%, 3.0%)

<0.001†

Positive Predictive Value

3.6%

(-2.8%, 9.9%)

0.269

-3.1%

(-5.8%, -0.5%)

0.019§/P >

Likelihood Ratio +

1.62

(-2.69, 5.92)

0.462

-1.59

(-3.68, 0.49)

0.134

Likelihood Ratio -

-0.09

(-0.18, -0.01)

0.021*

-0.08

(-0.12, -0.04)

<0.001†

Diagnostic Odds Ratio

14.3

(-8.06, 36.8)

0.210

3.63

(-4.99, 12.2)

0.409

* favours swabs taken at 36 weeks over 32 weeks

† favours either site over LVS only

§favours LVS only over either site

Hiller et al. BMC Pregnancy and Childbirth 2005 5:12   doi:10.1186/1471-2393-5-12