Sampling intervals of dynamic scintigraphy in lower gastrointestinal haemorrhage

The sampling rate for ^99mTc RBC scintigraphy in acute LGIH is quite variable. More rapid sampling allows more accurate localisation spatially by detecting bleeding more accurately temporally, minimising the impact of rapid peristalsis on localisation errors. There has been no report in the literature comparing the diagnostic efficiency of different sampling intervals. Methods: This was a retrospective clinical study of 49 ^99mTc red blood cell studies. Each was interpreted by four independent physicians as dynamic frames of 5 or 1 minute. Bleed detection, localisation and interpretive confidence was compared at the two rates. Results: Approximately 90% of bleed detection and interpretive confidence was derived from one minute sampling. Five minute sampling had greater effect on bleed localisation (28%). One minute sampling reduced false positive findings and five minute sampling reduced false negatives. Conclusion: One minute dynamic data should be reframed to five minute intervals to assist in bleed localisation. This is necessary only when a bleed has been identified on the one minute dynamic data.