P.E. Zollinger, R.S. Breederveld, W.E. Tuinebreijer, and R.W. Kreis reply:
Mr. Rogers and Mr. Ricketts are right that rats were used in the study by Kearns et al.1. However, the steady-state level of vitamin C in humans is mentioned in the discussion of their article. We wanted to combine the article by Kearns et al.1 with their citation of the earlier results reported by Levine et al.2, but it would have been better to mention them separately. Furthermore, we did not give the credit to Levine et al. that they deserve. We apologize for this.
Monitoring plasma levels is a very good suggestion, but it would have introduced an invasive component to our study, which might have dissuaded the volunteers from participating. It is difficult to take blood samples for the analysis of placebo blood cell counts. The intention-to-treat principle stands even if patients do not take their trial medicine, and these results have to be valued as input to the study as well. We believe that noncompliance results in an underestimation of the vitamin-C effect.
The criteria with which to diagnose complex regional pain syndrome are under debate all over the world, but no longer in The Netherlands. Here, a multidisciplinary scientific committee3 chose the criteria of Veldman et al.4 (and for research purposes, those of the IASP and Bruehl et al.5 in addition). This means that the studies and results of Atkins et al.6,7 are not taken into account here. We appreciate those authors' liberal way of diagnosing complex regional pain syndrome, and this would have given even more strength to our own study, but the trend in The Netherlands is to use the criteria reported by Veldman et al.4, which are based on the world's largest described population (n = 829).
In our study, the actual diagnosis of complex regional pain syndrome was not made over the telephone but rather in the outpatient clinic according to the described scheme of planned controls. The telephone inquiry after one year was to make sure that not one patient with complex regional pain syndrome type I had escaped our attention. The telephone inquiry provided no indication that any patient had been missed; no patient had to be called back for additional physical examination (but patients would have been called back if necessary).
Randomization ensures equality of the groups being compared. Confounding factors might exist, but adequacy of fracture reduction is not one of them according to this and our earlier study in The Lancet8. This point has been mentioned in our previous letter9. A difference in compliance is ruled out by the randomization as well.
Once again, we must state that our study was not an investigation of the effect of physiotherapy in patients with complex regional pain syndrome, but was a search for the incidence of and probable prophylaxis for complex regional pain syndrome in patients with a traumatic wrist fracture.
These letters originally appeared, in slightly different form, on . They are still available on the web site in conjunction with the article to which they refer.