Skip to main content

Table 2 Test characteristics of microvolt T-wave alternans as a predictor of mortality or severe arrhythmias

From: Microvolt T-wave alternans as a predictor of mortality and severe arrhythmias in patients with left-ventricular dysfunction: a systematic review and meta-analysis

  

Patients with Endpoint (n/N)

Non-negative vs Negative MTWA†

Study

Endpoint*

Positive

MTWA

Negative

MTWA

Indeterminate MTWA

Sensitivity

Specificity

PPV

NPV

Chow 2006 [18]

ACM

78/514††

21/254

NR

87.6

34.8

15.2

91.7

 

AD

33/514††

9/254

NR

78.6

33.7

6.4

96.5

Bloomfield 2006 [6]

ACM/SA/ICDS

47/360††

4/189

NR

92.2

37.1

13.1

97.9

Grimm 2003 [19]

CD/SA

18/137

7/72

13/54

82.6

28.9

16.2

90.3

Klingenheben 2000 [22]

CD/VT/VF

11/52

0/33

2/22

100

35.1

17.6

100

Ikeda 2000 [20]

SA

14/50

1/52

NR

NE

NE

NE

98.1

Kitamura 2002 [21]

CD

3/46

0/37

NR

NE

NE

NE

100

 

SA

8/46

1/37

NR

NE

NE

NE

97.2

 

CD/SA

11/46

1/37

NR

NE

NE

NE

97.2

Sarzi Braga 2004 [24]

CD

7/23

0/13

0/8

100

35.1

22.6

100

Sakabe 2001 [23]

SA

13/24

0/6

NR

NE

NE

NE

100

  1. Abbreviations: ACM: all-cause mortality; AD: arrhythmic death; CA: cardiac arrest; CD: cardiac death; ICDS: implantable cardioverter defibrillator shock; NPV: negative predictive value; NE: not estimable; NR: not reported; PPV: positive predictive value; SA: severe arrhythmias (ventricular fibrillation or ventricular tachycardia); VF: ventricular fibrillation; VT: ventricular tachycardia;
  2. * The inclusion of more than one clinical endpoint per row represents the use of a composite endpoint.
  3. † Test characteristics are for non-negative (i.e., positive and indeterminate) vs negative MTWA. Valid estimates of the sensitivity, specificity, and PPV can only be obtained when indeterminate tests are included. Consequently, test characteristics for Grimm [19], Klingenheben [22], and Sarzi Braga [24] were recalculated using non-negative or negative MTWA. Sensitivity, specificity, and PPV for studies that excluded indeterminate tests were not estimable (NE).
  4. †† The studies conducted by Chow [18] and Bloomfield [6] categorized participants as non-negative or negative MTWA. Consequently, these data represent patients with endpoints who had a non-negative test.