The efficacy of physiotherapy was assessed in a prospective study by comparing two groups of patients who were randomly assigned to a physiotherapy group (exercise therapy) or to a control group. The study was performed from July 2003 to January 2005 and patients enrolled between August 2003 and June 2004. Patients were recruited from the Radboud University Nijmegen Medical Centre (RUNMC) and the Canisius Wilhelmina Hospital (CWZ) Nijmegen in The Netherlands. Patients with breast cancer having to undergo surgery with ALND, were considered eligible for the study and had to meet the following criteria: 18 years of age and older with an ALND, following breast cancer, a Visual Analogue Scale (VAS; 0–10) pain score of 1 or more and moderate shoulder disabilities in daily life (minimal 3 points on a 5 points disability score list). Patients were excluded with a previous contra-lateral breast cancer surgery and insufficient knowledge of the Dutch language to fill in the questionnaires. Informed consent was obtained from all subjects, and the regional medical ethics board approved the study.
Patients were given a subject information sheet by a member of the nursing staff during their hospital stay. Patients who were willing to participate in the trial attended the Department of Physiotherapy for physical assessments two weeks after surgery, which was concurrent to the first outpatient clinic visit to the surgeon. Baseline measurements were assessed and patients who met the inclusion criteria signed an informed consent. Random assignment was done by an independent co-worker of the department into one of the two groups. The treatment group received specific physiotherapy treatment and the control group had no physiotherapy. Concealed randomization was achieved using a computer-generated random list, which was kept by the co-worker. All assessments were done at the RUMC Department of Physiotherapy by a single researcher, who did not participate in the treatment of the patients. The researcher was blinded to the treatment allocation and patients were instructed not to discuss their treatment with the researcher. The researcher made a note after the final assessment, to which group allocation that she thought the patient belonged.
Patients assigned to the control group received a leaflet flyer with advice and exercises for the arm/shoulder for the first weeks following surgery and had no further contact with a physiotherapist.
Patients assigned to the treatment group started physiotherapy two weeks following surgery in a private practice of their own choice. The research assistant contacted the individual physiotherapists (n = 15) who had agreed to comply with the treatment regime and supplied them with information regarding the project and treatment guidelines. This information consisted of:
- guidelines with advice and exercises for arm/shoulder, posture correction, coordination exercises, exercises for muscular strength and improvement of general physical condition ;
- exercises to prevent lymph edema ;
- instruction for soft tissue massage of the surgical scar if required;
- a form to report the content of the treatment sessions and a 3-point scale to indicate whether the amount of treatment sessions was sufficient.
The total number of treatments was nine (nine being usually covered by the healthcare insurance), once or twice weekly for the first three weeks, and thereafter once a fortnight or less. The total amount of sessions had to be given within three months. Patients were asked to perform home exercises for ten minutes each day.
Demographic data was recorded (age, general health) and as well as data and information about the level of impairment, disability and participation at baseline and after three and six months in both groups.
The primary outcome variables were pain in the shoulder/arm, measured using the VAS score (0 – 10, 0 = no pain; 10 = unbearable pain) and shoulder mobility (flexion [0–180°], abduction [0–180°]), measured by use of a digital inclinometer under standardized conditions.
Secondary outcome measures were disabilities in daily life, measured by the DASH (Disabilities of the Arm, Shoulder and Hand) questionnaire  (0 – 100, 0 = no functional problems, 100 = maximal problems), edema (ml), measured in both arms by means of water displacement, grip strength (Kg) of both hands, measured using the hand-held dynamometer and quality of life, as measured by the SIP (Sickness Impact Profile-short version) questionnaire (0 – 68, 0 = good health status; 68 = severe physically disabled) . The total amount of time for each measurement session was approximately 40 minutes, measurements taking place prior to randomization at intake and at three and six months following intake.
Data was analyzed using the SPSS version 12.1. Univariate analysis of variance was used to test differences in outcome variables between the control group and physiotherapy group. Baseline data were entered in the analysis as co-variates. Level of significance was set at 0.05.