We thank Dr Tan for his interest in our recent publication.1 We agree that variation in estimates of the prevalence of pedal edema in existing studies may result from variability in the ascertainment of edema, differences between study populations (in propensity to develop pedal edema in response to dopamine agonists and in background risk of edema due to other causes), and chance variation, particularly when small populations are studied. For example, Dr Tan reports an incidence of edema of 15% in his study population of 20 patients; the 95% confidence interval around this estimate is 3% to 39%! Methods for ascertaining whether variations between studies are due to random chance or true heterogeneity are described in the systematic review literature.2 As we note in our article, our estimate of prevalence of edema is subject to all the limitations that attend a retrospective medical-record review conducted in a specialty clinic but is consistent with that seen in the prospective, placebo-controlled CALM-PD randomized trial,3 which likely represents the most credible single estimate to date of edema risk in patients taking pramipexole.
Kleiner-Fisman G, Fisman DN. Peripheral Edema and Dopamine Agonists in Parkinson Disease—Reply. Arch Neurol. 2007;64(10):1547. doi:10.1001/archneur.64.10.1547