The review included 17 studies, 14 of which were prospective, including five randomised controlled trials (RCTs), and three were retrospective. The sample sizes ranged from 56 to 4,160.
In 12 studies reporting pain, the prevalence ranged from 8% to 36% one year after treatment and from 7.5% to 21.2% two years after treatment. Range of motion was reported in 10 studies and the prevalence of restricted motion in the affected arm ranged from 6.4% to 31% one year after treatment and from 0.0 to 9.4% two years after treatment. Oedema was reported in 13 studies and the prevalence of arm oedema ranged from 3.0% to 10.0% at six months and from 6.0% to 14.0% at one year or later. Strength was reported in three studies and the prevalence of strength reduction ranged from 17.0% to 19.0% at one year or later; the text stated that six studies assessed strength, but the results were only reported for three studies in the table. Sensory disorders were described in all 17 studies and prevalence was 2.0% to 66.0% at one year or later, remaining at 1.0% to 22.0% at two years.
There was strong evidence for the predictive value of time after surgery and young age for late morbidity; moderate evidence for the predictive value of breast surgery, axillary radiation, and tumour location; and weak evidence for the predictive value of a two-step procedure, body mass index, and the use of radiotherapy pharmaceutical alone.