We thank Brunetti and colleagues for their comments. They make some interesting points about the potential biases inherent in crossover trials, but we do not feel that they invalidate the conclusions of our study showing the effectiveness of quinine as a treatment for nocturnal leg cramps. The initial "run-in" period was to ensure some measurable cramp activity and to act as a washout for any quinine or vitamin E that patients may have been taking on their own. Since we did not use a parallel-group design, determining a baseline was not necessary to ensure similarity between treatment groups. Our goal was to compare quinine and vitamin E with placebo, so we used these drug-on periods for our data analysis rather than the run-in period. Comparing time periods 1, 2, and 3 showed no significant difference, indicating the lack of a period effect.
We also found no evidence of a carryover