A venous blood sample of 8.5 ml were obtained from patients. For the study group, the blood samples were mixed with 1.5 ml ACD-A or sodium citrate to achieve anticoagulation. Resulting material was then centrifugated for 5 minutes with RCF 1200 G speed to clump erythrocytes, and then centrifugated for 10 minutes in same speed to obtain thrombocyte concentrate. 2 ml's of the platelet rich plasma was injected to the shoulder of the subjects. For the sham injection group, same amount of blood was taken and they were given a of waiting time of same duration with the study group, with the resulting preparate being 2 ml's of 0,9% saline instead. PRP solutions were prepared with kits of Easy PRP(Neotec Biotechnology, Istanbul, Turkey).

Injections were done every two weeks, for a total of 3 times.

Injections containing saline were done every two weeks, for a total of 3 times. 21 G injection needles with injection technique of posterior approach were used. To provide blindness, all injections were done using injectors coated with non transparent tape. All of the PRP injections were done by the same physician each time, with compliance to preventive measures against complications such as infections. Patients and the physician who applied the injection were blinded to the groups, and the solution was prepared by another researcher who was not blind to the groups.

A venous blood sample of 8.5 ml were obtained from patients. For the study group, the blood samples were mixed with 1.5 ml ACD-A or sodium citrate to achieve anticoagulation. Resulting material was then centrifugated for 5 minutes with RCF 1200 G speed to clump erythrocytes, and then centrifugated for 10 minutes in same speed to obtain thrombocyte concentrate. 2 ml's of the platelet rich plasma was injected to the shoulder of the subjects. For the sham injection group, same amount of blood was taken and they were given a of waiting time of same duration with the study group, with the resulting preparate being 2 ml's of 0,9% saline instead. PRP solutions were prepared with kits of Easy PRP(Neotec Biotechnology, Istanbul, Turkey).

Injections were done every two weeks, for a total of 3 times. 21 G injection needles with injection technique of posterior approach were used(18, 19). To provide blindness, all injections were done using injectors coated with non transparent tape.

Visual Analogous Scales for pain were used as outcome measures. Both resting and activity pain levels were evaluated. Patients were asked to describe their pain, level of disability or treatment satisfaction on a 10 cm long line, with 0 point being "none", and 10 cm point being "highest possible". The proximity of the marked point to the 0 end was measured, and the results were recorded

Shoulder range of motions were also evaluated as outcome measures. The measurements were done accordingly to the guidelines of American College of Orthopedic Surgeons by a conventional goniometer(20, 21). The researcher who performed the measurements was an experienced in this field, and was blinded to the treatment groups. External and internal rotation, adduction, abduction, flexion and extension of both shoulders were recorded for each group.