Aim: Knee arthroplasty is one of the best treatments to reduce pain and improve quality of life in cases of osteoarthritis resistant to osteoarthritis, however since it is an offensive act and delayed arthroplasty may lead to unwanted side effects after surgery. Therefore, the correct and timely selection of the patient for knee arthroplasty is important so it is necessary to collect standard and integrated data necessary. The purpose of this research is to identify minimum data requirements for predicting knee arthroplasty in patients with osteoarthritis.
Methods: This is a descriptive study conducted in 2017. Effective factors in knee arthroplasty prediction were identified Through library research and Internet search of keywords such as total knee arthroplasty, minimal data set, ... in valid databases such as PubMed, science direct, ... and the Google search engine Scholar. Sampling was done using Morgan table 32 faculty members of where rheumatology specialist at Tehran University of Medical Sciences were selected as sample. Validity of the questionnaire was assessed by Health Information Management Professors and its reliability was 82% Cronbach's alpha in SPSS 16 software. Results were analyzed using Excel software.
Results: The results of the two-stage Delphi technique include: nonclinical data (age, body mass index) and clinical data (rest pain, activity pain, quality of life, the x-ray findings, knee instability, knee deformity, and limited mobility). These are the Minimum data necessary for predicting the need of knee arthroplasty in osteoarthritis patients. 
Conclusion: The results of this study can be considered as the first step in the development of intelligent knee arthroplasty prediction systems. Using these systems will be very useful in managing the waiting list of surgery in health centers.