Shaken or Stirred? The Inconsistencies of Manual Contrast Media Dilution in Endovascular Interventions

MATERIALS AND METHODS: Phase I: Eleven radiological technologists were asked to fill a CM injector 3 times with 50% diluted CM (iopromide 300 mg I/mL). The dilution was injected (12 mL/s) through a Coriolis flowmeter, with CM concentration and total volume calculated. Interoperator, intraoperator, and intraprocedural variations were calculated as coefficients of variability. Contrast media dose reporting accuracy was determined. Phase II: The study was repeated after implementation of a standardized dilution protocol with 5 representative operators. RESULTS: Phase I: The average injected concentration among 11 operators was 68% ± 16% CM (n = 33; range, 43%-98%), as compared with the target of 50% CM. The interoperator variability was 16%, the intraoperator variability was 6% ± 3%, and the intraprocedural variability was 23% ± 19% (range, 5%-67%). This led to overdelivery of CM compared with intended patient dose by 36% on average. Phase II: After standardization, injections averaged 55% ± 4% CM (n = 15; range, 49%-62%), with interoperator variability of 8%, intraoperator variability of 5% ± 1%, and intraprocedural variability of 1.6% ± 0.5% (range, 0.4%-3.7%). CONCLUSIONS: Manual CM dilution can lead to substantial interoperator and intraoperator, as well as intraprocedural variability in injected concentration. This can result in underreporting of administered CM doses to patients. It is recommended that clinics assess their current standard of care regarding CM injections for endovascular interventions and evaluate potential corrective actions if appropriate.