Page 23 - Cardiac testing for coronary artery disease in potential kidney transplant recipients

Table 3 and Figure 6 summarise indirect comparison results.
Overall, there was evidence that DSE (13 studies) had better test cation and had reference thresholds ≥ 70% stenosis on coronary
accuracy than MPS (9 studies) (P = 0.02). Using the results from angiography were included in the analysis, there was no evidence
the earlier analysis, DSE appeared to have a higher pooled sensi- of a statistically significant difference between tests (P = 0.09).
tivity (DSE: 0.79 (95% CI 0.67 to 0.88) versus MPS: 0.74 (95% DSE (8 studies) appeared to have a higher pooled sensitivity: 0.78
CI 0.54 to 0.87) and specificity DSE: 0.89 (95% CI 0.81 to 0.94) (95% CI 0.59 to 0.89) than MPS 0.67 (95% CI 0.48 to 0.82) and
versus MPS: 0.70 (95% CI 0.51 to 0.84). The variability in ac- DSE specificity: 0.88 (95% CI 0.76 to 0.94) versus 0.77 (95% CI
curacy was smaller for DSE than MPS, demonstrated by the dif- 0.61 to 0.88)] compared with MPS (7 studies), as well as a higher
ference in size of the 95% confidence regions in HSROC space. corresponding AUC.
When we included only studies that used definitions of ≥ 70%
stenosis on coronary angiography to diagnose severe CAD, DSE Subgroup analyses
(9 studies) had pooled sensitivity and specificity of 0.76 (95% CI
Sparse data, both in terms of numbers of studies and study par-
0.60 to 0.87) and 0.88 (95% CI 0.78 to 0.94) respectively. MPS
ticipants, meant that we were unable to perform meaningful sub-
(7 studies) had pooled sensitivity and specificity of 0.67 (95% CI
group analyses on the effect of DM or prevalence of angina and
0.48 to 0.82) and 0.77 (95% CI 0.61 to 0.88) respectively. There
symptomatic ischaemic heart disease (IHD) on diagnostic test per-
was no statistically significant difference between tests (P = 0.09)
formance. Only one study (Vandenberg 1996) included a patient
(Figure 7). When we included only studies where partial verifi-
population who had no history of angina or IHD. Therefore, a
cation bias was avoided, DSE (10 studies) had pooled sensitivity
sensitivity analysis of diagnostic accuracy in studies that enrolled
and specificity of 0.80 (95% CI 0.64 to 0.90) and 0.89 (95% CI
only patients who had no symptoms of cardiac disease or history
0.79 to 0.95) respectively. MPS (8 studies) had pooled sensitivity
of IHD could not be conducted.
and specificity of 0.68 (95% CI 0.51 to 0.81) and 0.75 (95%
CI 0.60 to 0.86) respectively. The difference in accuracy between
MPS and DSE tests for these studies was statistically significant (P
Summary of results
= 0.03) (Figure 8). When only studies that avoided partial verifi-
Summary of results
Summary of results: Results of studies on cardiac testing in kidney transplant candidates

Review question: Comparison of non-invasive cardiac screening tests with coronary angiography for the detection of significant CAD
in potential kidney transplant recipients
Patient population: Kidney transplant candidates undergoing pre-transplant cardiac evaluation
Setting: Investigations performed in hospital or in an outpatient setting
Geographical location: Studies were conducted in USA (12 studies), Brazil (4 studies), India, (3 studies) the UK (3 studies), Australia
(1 study), Canada (1 study), and Spain (1 study)
Index test : Any non- or minimally invasive test used to assess risk of CAD
Reference standard: Coronary angiography
Included studies: DSE (13 studies; 745 participants), MPS (9 studies; 582 participants), EST (2 studies; 129 participants), EBCT
(1 study; 97 participants), DSF (1 study; 86 participants), exercise ventriculography (1 study; 35 participants), CIMT (1 study;
105 participants), resting wall motion abnormality on echocardiography (2 studies; 265 participants), left ventricular dysfunction
on echocardiography (1 study; 52 participants), mitral annular calcification on echocardiography (1 study; 125 participants), resting
ECG (3 studies; 263 participants)

Limitations

Only DSE and MPS were evaluated in detail, although these also had only a limited number of included comparisons with small
sample sizes. No studies were found investigating cardiopulmonary exercise testing, CT coronary angiography, magnetic resonance
angiography or cardiac magnetic resonance imaging. Fewer than five studies were found for each of EBCT, resting ECG, conventional
echocardiography, exercise ventriculography, DSF and CIMT. Sparse directly comparative data also resulted in low power to detect
important differences in accuracy between tests
Significant heterogeneity was present among studies investigating the same screening test. Although differences in study population

Cardiac testing for coronary artery disease in potential kidney transplant recipients (Review) 21
Copyright © 2011 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.