2018 Section 6 - Laryngology, Voice Disorders, and Bronchoesophalogy

Fig. 1. Study selection flowchart. [Color figure can be viewed in the online issue, which is available at www.laryngoscope.com.]

analysis is presented in Figure 5. The size of the bubbles reflects the relative size of each study. The summary estimate of sensitivity and specificity is depicted in the quadrant, and the 95% confidence and prediction regions are represented by dotted lines. Results of Wilson’s scores are presented in Figures 3 and 4. The sensitivities of VS within the single studies ranged from 86% to 100%, and corresponding specific- ities ranged from 7% to 93%. The meta-analysis showed that the sensitivity of the combined results was 0.96 (95% CI: 0.89-0.98), and the specificity was 0.65 (95% CI: 0.21-0.93) (Fig. 5). This means that the test is gener- ally highly sensitive and thus able to identify the

Risk of Bias Assessment In line with SIGN recommendations, five domains were assessed for each of the five included studies. Thus, a total of 25 assessments were carried out. Twelve of these were high quality, and the remaining 13 were acceptable quality. Overall assessment was high quality for three of five studies, and acceptable quality for two studies. Assessments are summarized in Table II.

Synthesis of Results A meta-analysis was conducted on five studies with study populations ranging from 32 to 130 patients, giv- ing a total of 307 patients. The bubble plot of the meta-

TABLE I. Variables Extracted From Included Studies.

n

TP

FP

FN

TN

Sensitivity (CI)

Specificity (CI)

PPV

NPV

Peretti et al. 18

42

14

3

1

24

0.93 (0.70–0.99)

0.89 (0.72–0.96)

0.82

0.96

Gugatschka et al. 11 El-Demerdash et al. 17

130

32

77

0

21

1 (0.89–1.00)

0.21 (0.14–0.31)

0.29

1

60 43

31 14

2

1 1

26

0.97 (0.84–0.99) 0.93 (0.70–0.99)

0.93 (0.77–0.98) 0.07 (0.02–0.23)

0.94 0.35

0.96 0.67

Colden et al. 7 Caffier et al. 12

26

2

32

12

2

2

16

0.86 (0.60–0.96)

0.89 (0.67–0.97)

0.86

0.89

CI 5 confidence interval; FN 5 false negative; FP 5 false positive; n 5 patients included; NPV 5 negative predictive value; PPV 5 positive predictive value; TN 5 true negative; TP 5 true positive.

Laryngoscope 126: September 2016

Mehlum et al.: VS and Prediction of Early Glottic Cancer

23