BackgroundUsing myocardial perfusion scintigraphy (MPS), an increase in left ventricular (LV) volumes or a decrease in ejection fraction (EF) from rest to stress may be clinically important. The variation in these measures between the low-dose stress acquisition and high-dose rest acquisition in a one-day stress-rest protocol has not been established. We assessed the reproducibility of gated volumetric indices between stress and rest and the normal variation in ungated TID ratio for a one-day stress-rest 99mTc-tetrofosmin protocol. MethodsTwo thousand and one hundred and fifty eight (2158) 99mTc-tetrofosmin MPS patient studies were analyzed retrospectively. Studies were excluded for incomplete data, significant technical difficulties, or (for gated analysis but not for analysis of TID ratio) if the LV EF was > 75%. An analysis of gated data was undertaken to establish the reproducibility of ventricular volumes and EF between stress and rest scans. Ungated volume data were analyzed to determine the confidence limits of TID ratio according to ventricular volume. ResultsGated data were analyzed for 621 patients without inducible hypoperfusion. Mean EF at rest was slightly higher than after stress (62.4% ± 10.3% vs 61.2% ± 10.4%, P < 0.001), and the standard deviation of the difference was 5.2% (95% CI 4.9% to 5.5%). Ungated volumes were available for 992 non-ischaemic patients. The upper 95% CI for TID ratio was 1.23. This increased from 1.20 to 1.37 between the highest and lowest deciles of rest ungated volume. ConclusionUsing a one-day stress-rest 99mTc-tetrofosmin protocol, a fall in LV EF between rest and stress of > 11.6% or a TID ratio of > 1.23 is likely to be clinically reliable. The upper limit of normal for TID ratio needs to be increased for patients with small LV chamber volumes.