Despite the presence of established treatment guidelines, chronic pain remains an undertreated condition (Weinstein et al 2000 ; Glajchen 2001 ). Almost half of all patients in routine clinical practice experience inadequate relief from chronic pain (Cherny and Portenoy 1994 ; Glajchen 2001 ). Low patient satisfaction with analgesic therapy is further evidenced by the fact that Americans over age 60 years consult at least 3 physicians about their pain medications (APF 2006). There are a variety of factors leading to the low effectiveness of chronic pain therapy, including patient factors, physician education, regulatory oversight, and formulary issues. The relatively low effectiveness of chronic pain therapy may be related to various limitations of currently used analgesics (Table 2 ), as well as conservative dosing by physicians (ACR 2000 ; Weinstein et al 2000 ; Stephens et al 2003 ; Hudson et al 2005 ; Caldwell et al 2006 ). Treatment with NSAIDs is restricted by their association with increased risk for GI side effects, renal toxicity, and cardiovascular toxicity, especially in the elderly (ACR 2000 ). Risk factors for reversible renal failure in patients with intrinsic renal disease receiving nonselective NSAIDs include age (≥65 years), hypertension, and congestive heart failure (Page and Henry 2000 ). Studies have shown that subjects with a history of heart disease receiving nonselective NSAIDs had a 10.5-fold increased risk for congestive heart failure, compared with non-NSAID users (Page and Henry 2000 ). Recent studies have shown that COX-2 inhibitors are similarly associated with cardiovascular safety issues. Using published and unpublished data between 1966 and 2005, a meta-analysis of randomized comparative trials of at least 4 weeks’duration showed both selective COX-2 inhibitors and certain high-dose nonselective NSAIDs significantly increased the risk for cardiovascular events (Kearney et al 2006 ). Compared with placebo, COX-2 inhibitors showed a relative increase of 42% in serious vascular events (p = 0.003), which was similar to that found with certain nonselective NSAIDs. (Kearney et al 2006 ).