OBJECTIVETo observe the effects between acupuncture combined with rehabilitation and simple rehabilitation for shoulder pain after stroke, and to select the best plan of acupuncture manipulations at different stages by orthogonal design.

METHODSNinety patients were treated with comprehensive rehabilitation, and nine cases without acupuncture were arranged into a control group. Eighty-one patients of orthogonal design were applied by acupuncture with the same acupoints and course. The VAS score and its weighted value were regarded as the observation indices,and the effects between the acupuncture group and the control group were compared. The optimal plans of acupuncture manipulations of the early stage and the later stage were chosen after the first course treatment and the third course treatment separately. The acupuncture depth (factor A:A: shallow depth less than 25 mm, A(II): modest depth 25-40 mm, A(III): deep depth 40-50 mm), the acupuncture angle (factor B:B(I): perpendicular insertion, B(II): horizontal insertion, B(III): oblique insertion), needle manipulated frequency (factor C: C(I): zero time, C(II): one time, C(III): three times) and needle retained time(factor D:D(I):20 min, D(II): 30 min, D(III): 60 min) were studied. The differences among all factors and the diversity among major factors at different stages were analyzed.

RESULTS(1) Acupuncture combined with rehabilitation at the early and the later stage acquired better improvement than simple rehabilitation (all P < 0.01). (2) The optimal acupuncture manipulation plan at the early stage was A(III) B(III) C(I) D(I), which was deep acupuncture and oblique insertion for 20 min with zero-time manipulation; the optimal acupuncture manipulation plan at the later stage was A(III) B(III) C(III) D(I), which was deep acupuncture and oblique insertion for 20 min with three-time manipulation. (3) There was significance for acupuncture depth and angle at the early stage (both P < 0.01) and there was significance for insertion depth, acupuncture angle and manipulating frequency at the later stage (all P < 0.05). (4) At the early stage, the insertion depth was statistically significant between A(I) and A(II), A(I) and A(III), A(II) and A(III) (P < 0.05, P < 0.01), and the statistical significance was existed between B(I) and B(III) (P < 0.01). At the later stage, the insertion depth was statistically significant between A(I) and A(III), A(III) and A(II), A(I) and A(II) (P < 0.05, P < 0.01), and the statistical significance was existed between C(I) and C(III), C(II) and C(III) (P < 0.05).

CONCLUSIONAcupuncture combined with rehabilitation acquire apparent effect for shoulder pain after stroke. At the early stage,the optimal plan is deep and oblique insertion for 20 min with zero-time manipulation. At the later stage, the best plan is deep and oblique insertion for 20 min with 3-time manipulation.