benefits and safety of each nonpharmacologic modality reviewed. Again, initial voting was done privately with votes submitted electronically using Excel spreadsheets. If consensus was not achieved with private voting, further discussion of the modality was conducted at the meeting with open voting and group discussion until consensus was achieved. Throughout the voting process, if one or more members of the TEP reported a conflict of interest concerning any specific modality, they were encouraged to recuse themselves from the discussion and voting on that modality. The summary data from the systematic literature review that were provided to the TEP before its meetings are available as supplementary files in the online version of this article at -4658. These data include not only the citations for the most relevant systematic reviews and/or randomized clinical trials and how to interpret them (in the OA guideline development process document), but also the summary of findings tables indicating the efficacy and safety/tolerability of pharmacologic modalities for hip and knee OA, duloxetine, nonpharmacologic modalities for hip and knee OA, and updated tai chi. In addition, the top-line citations for each treatment modality that were used for the development of the summary of findings tables are included in Supplementary Appendix B The final set of recommendations was drafted after discussion of the evidence at each TEP meeting. Consensus was defined as 75% or more of the members of the TEP voting to either strongly or conditionally recommend using a modality, either strongly or conditionally recommend not using a modality, or choosing not to make a recommendation on the use of a modality (23). A strong recommendation for using a modality required high-quality evidence and evidence of a large gradient of difference between desirable and undesirable effects of the treatment (i.e., benefits compared to harms). A conditional recommendation for using a modality was based on absence of high-quality evidence and/or evidence of only a small gradient of difference between desirable and undesirable effects of the treatment. In addition, when there was more uncertainty and/or variability in the values and preferences of the TEP members for a specific modality, this more likely resulted in a conditional recommendation. The lack of data from appropriate RCTs resulted in either not making a recommendation or making a recommendation not to use a modality, depending on the harms of the modality in other conditions and/or the values and preferences of TEP members. The recommendations of the TEP focus on the initiation of treatments for OA of the hand, hip, and knee. Costs of care were not considered in formulating these recommendations. The initial literature searches were updated during the first quarter of 2011 with a cutoff date of December 31, 2010 by the ACR staff using the identical methodology and search filters as in the original searches. The results of the updated literature search of pharmacologic agents were reviewed by physician members of the project's Steering Committee (RDA, MCH, TT, PT) to identify studies that might provide information on new treatments or new information on the efficacy or safety/tolerability of existing treatments. The results of the updated literature search for nonpharmacologic modalities were reviewed by the rheumatology health professionals who were members of a TEP to identify studies that might