This article from Switzerland is one of the first that shows poor results with 16-slice CT, in unselected patients..
The results shown in this article, in 149 consecutive patients, are really dismal with an overall sensitivity of picking up significant CAD of 86%, but a specificity of just 49%. On a per segment basis, the sensitivity was just 30%.
This article obviously needs to be read in more detail. What is seen, is that the patients were truly unselected no calcium scoring was performed before the procedure. No attempt was also made to control the heart rate and though not statistically significant, those with higher heart rates had poorer results.
The biggest problem according to the authors was calcification, which both masked and simulated disease, where it was present. A similar article by Achenbach's group last year in a smaller series of 33 patients, came to almost similar conclusions.
This is of interest because there is now a trend with 64-slice CTs, to disregard the presence of calcification and to proceed with CT angiography. As the cases (case 1, case 2, case 3, case 4) in the accompanying images blog show, as long as there is no calcification and the attention to technique is properly maintained, the results are excellent. But the moment there is calcification, things really start going wrong and we now report often that "dense calcification is seen preventing adequate lumen visualization". In my opinion, it is often easier to common on in-stent restenosis, than to see the lumen through dense calcium.
Link: Limited diagnostic yield of non-invasive coronary angiography by 16-slice multi-detector spiral computed tomography in routine patients referred for evaluation of coronary artery disease -- Kaiser et al. 26 (19): 1987 -- European Heart Journal.