How common is complex regional pain syndrome-Type I?

Despite the regularity with which CRPS-I is seen in pain clinics, epidemiological data on its occurrence in the general population have been sparse. In part due to historical disagreements regarding mechanisms and diagnosis of CRPS-I, clinical beliefs regarding its incidence and prevalence range widely. While some argue that CRPS-I does not even exist as a neuropathic pain disorder (Ochoa, 2006), clinicians receiving many CRPS referrals assume a much higher rate of occurrence. The more rare CRPS is believed to be, the less likely it is to be considered a relevant diagnostic rule-out, particularly among physicians not specializing in pain. Patient outcomes may suffer if appropriate treatment is delayed (Stanton-Hicks et al., 2002). Sandroni et al. (2003) published the first population study of CRPS-I. Responses from both researchers (e.g., Bennett and Harden, 2003) and the CRPS patient community indicate that conclusions of this study were somewhat controversial. Statements that CRPS-I is rare (5.46 new cases per 100,000 annually) and associated with frequent ''spontaneous resolution'' provoked strong reactions. As noted by Bennett and Harden (2003), conclusions that ''spontaneous resolution'' of CRPS-I is common were unjustified because over 90% of the sample received physical therapy, and nearly half received sympathetic blocks and pharmacological intervention. Until now, however, no other epidemiological data were available to support or refute the reported low incidence of CRPS. The article by de Mos et al., 2007 (this issue) is only the second published epidemiological study regarding CRPS incidence in the general population. When based on clinical diagnoses confirmed by the original treating physicians, the incidence was 26.2 new cases per 100,000 annually, a figure 4.2 times higher than the Sandroni et al. study. Even when restricted to those cases in which detailed specialist evaluation data were available to make independent diagnoses using IASP diagnostic criteria, de Mos et al. report an incidence of 16.8 new cases per 100,000 annually, nearly 3 times higher than Sandroni et al. These significant discrepancies in CRPS incidence demonstrate the importance of continued epidemiological investigations. Although both studies are valuable, we believe that certain features of the de Mos study may have produced relatively more accurate incidence estimates. Compared to Sandroni et al. strengths of this new study include a study population more than twice as large (217,653 versus 106,470) and the fact that the study period began after publication of the 1994 IASP diagnostic criteria so was less likely to be influenced …