GRADE assessment of evidence quality of urine dipstick (nitrites) accuracy (index test) to detect asymptomatic bacteriuria (reference standard: urine culture) in pregnancy8
| Outcome (presence or absence of bacteriuria) | No of studies (No of patients) | Accuracy measures | Features that may affect evidence quality* | Quality of evidence‡ | ||||
| Risk of bias | Indirectness | Inconsistency | Imprecision | Publication bias† | ||||
| True positives (patients with asymptomatic bacteriuria) | 21 cross-sectional studies (699 patients) | Sensitivity 0.56 (95% CI 0.42 to 0.69) | Serious§ | Serious¶ | Serious** | Serious†† | – | ⨁◯◯◯ Very low |
| False negatives (patients incorrectly classified as not having asymptomatic bacteriuria) | ||||||||
| True negatives (patients without asymptomatic bacteriuria) | 21 cross-sectional studies (8560 patients) | Specificity 0.99 (95% CI 0.98 to 0.99) | Serious§ | Serious¶ | Not serious | Not serious | – | ⨁⨁◯◯ Low |
| False positives (patients incorrectly classified as having asymptomatic bacteriuria) | ||||||||
*As specified in table 1.
†Domain was not assessed (see text for justification).
‡See figure 1 for graphic display of quality of evidence.
§57% of included studies of low and moderate quality14 (in the cited reference the detail quality assessment appears in appendix 5).
¶54% of included studies were assessed as ‘high’ or ‘unclear’ concern over applicability.
**Visible variability between studies on the forest plot.
††Noticeable imprecision, wide CIs.