This is the first survey to investigate and describe physiotherapy-supervised mobilization and exercise after cardiac surgery in Sweden. Most of the physiotherapists, in total 90%, declared that they routinely met all patients undergoing cardiac surgery, while 10% responded that they only treated certain patients, with special indications or special needs. The physiotherapy treatment was most frequently given on the first two postoperative days. On day 1 the patients usually received one to three treatment sessions by the physiotherapist, and on day 2, they were given one to two treatment sessions. The main purpose of physiotherapy after cardiac surgery was mostly seen as preventing and treating postoperative complications, improving pulmonary function and encouraging physical activity. Written local physiotherapy guidelines or protocols for physiotherapy management of cardiac surgery patients were available, according to 21 out of the 29 respondents.
Only one previous survey of physiotherapy management of patients undergoing cardiac surgery has been found, performed by Tucker et al.  in Australia and New Zealand. To our knowledge, our study is the first European survey describing physiotherapy treatment after cardiac surgery.
The clinical practice in Sweden and Australia and New Zealand seems to be similar in terms of the components of postoperative physiotherapy treatment, assessment of physiotherapy given to all patients (89%), and mobilization and breathing exercises, as described by Tucker et al. . However, the study was carried out in 1996, so we do not know how their clinical routines and practice come across and may differ today. More recently physiotherapy management after thoracic surgery was described in a survey study by Reeve et al. , however the physiotherapy treatment following thoracotomy cannot be compared to treatment after cardiac surgery.
In total, 29 replies were received out of the 33 questionnaires sent out. Since the questionnaires were comprehensive the response rate of 88% can be considered high. A high response rate is important and various strategies were used to improve the response rate. Comprehensible instructions were given, the questionnaires were printed on coloured paper; stamped, addressed envelopes were included with the questionnaires and reminders were sent out where the questionnaires had not been returned.
Access to a list of all physiotherapists working in departments of cardiothoracic surgery as well as personal contacts with physiotherapists at all departments ensured that all relevant physiotherapists were included in the survey. The study of a total population sample and the high response rate gives the study good external validity. It is likely that the results of this survey reflect current practice in Sweden, even if some important questions may have been overlooked and the exact description of the actual clinical practice, in observational studies, is warranted in the future.
An intrinsic selection bias in questionnaire studies is a risk if only the most motivated physiotherapists respond. Since only four physiotherapists failed to answer, we found this risk of bias fairly low. Because no nationally developed questionnaire for this purpose existed, the authors designed the questionnaire. To improve the content validity of the survey, information from earlier questionnaires used in similar studies [8, 9] as well as pilot testing was used to construct the questionnaire. Despite these limitations we believe that the results from this survey provides a good overview of the physiotherapy treatment given to cardiac surgery patients.
The majority (90%) of the physiotherapists offered preoperative information to all patients undergoing non-emergency cardiac surgery, which is similar (94%) to the routines in Australia and New Zealand described by Tucker at al. . The educational content of the preoperative information was similar, with early mobilization, post-sternotomy recovery and postoperative pulmonary function being the topics most covered.
Treatment was generally less comprehensive during weekends. Routine physiotherapy for patients on their first postoperative day was given more often on Saturdays (59%) than on Sundays (31%). For patients on their second postoperative day, no routine physiotherapy was given on Sundays, except where needed, as reported by half of the physiotherapists. These results indicates that there is a discrepancy in treatment of patients depending on which weekday they are operated on in Sweden. By comparison, in Australia and New Zealand during the 1990's, evening services were provided as required in 71% of hospitals, while in Sweden no evening physiotherapy treatment is available at all.
In the late 1960 s, patients would spend at least 3 weeks resting in bed after cardiac surgery. Since then the practice of postoperative physiotherapy has changed in response to advances in medical and surgical knowledge . Today there is an agreement as to the value of early mobilization and positioning after cardiac surgery [11–13], despite the risk of postoperative cardiac dysfunction [6, 14]. Almost all physiotherapists in our study mobilized their patients with regard to sitting and standing on postoperative day 1. Invasive cardiovascular monitoring is common in the early postoperative period and affects the ability to walk a longer distance from the bed because of the equipment.
Of course, it is the individual strength and cardiovascular status of the patient that decides the level and intensity of mobilization. In this study the average mobilization routines performed by a physiotherapist of a hypothetical "everyday" patient was determined. The actual mobilization of individual patients has not been the focus of the present study. Despite the frequent use of early mobilization, the benefit of mobilization in preventing postoperative complications has not been studied in the cardiac surgery patient. Studies' investigating different levels of mobilization during the hospital stay are lacking. In a recent follow-up of CABG patients, work capacity, and participation in household work were described as predictors of continuation at work after the surgery . The authors encouraged medical personnel to activate the cardiac surgery patient to undertake household work and all kinds of physical activities .
By contrast, positioning to a side-lying was used only by approximately 25% of the physiotherapists during the first postoperative days, despite the fact that positive effects of side lying on lung volumes  and oxygenation  have been described. Patients possibly experience increased pain and discomfor in this position, which may be an explanation for the low frequency of use.
All physiotherapists provided information about physical activity, exercises and rehabilitation to patients after discharge from the hospital. The content of the information would be interesting to study further, as could recommendations and regimens from cardiac surgeons, anaesthesiologists and cardiologists.
Shoulder range of motion exercises are today a common form of therapy intended to improve ventilation, preserve thorax mobility and ease sternal circulation and healing , even though the efficacy of shoulder range of motion exercises has been questioned .
Instructions in range of motion exercises for the upper extremities and thorax were mostly started on postoperative days 2 and 3. Only six of the physiotherapists started these exercises on the first postoperative day. It is currently not known how these exercises should be performed. In a study of patients with chronic sternal instability, by El-Ansary et al. , it was shown that bilateral upper limb movements were significantly less associated with sternal pain compared with unilateral movements. In the present survey, mostly bilateral upper extremity exercises (69%) were prescribed, rather than unilateral range of motion exercises. How many times the patients were instructed to perform the exercises varied between one and three times a day. Shoulder range of motion exercises, to be continued after discharge, were given by all physiotherapists. Recommendations for continuing the exercise programme varied between 1 and 8 weeks, however.
Recommendations for sternal precautions during the first postoperative weeks differed, which may reflect differences between recommendations from thoracic surgeons and hospital policy. Diverse instructions were given regarding restrictions of using of arms to push up from a lying to a sitting position, using the stomach muscles and also using crutches. However, almost all of the physiotherapists allowed the patients to use their arms to push up from sitting to standing position, move their arms and shoulders in full active movement, and use rolling walkers and walkers. Instructions for moving in and out of bed were given to the patients using a "standard technique" by 90% of the physiotherapists. The most commonly described technique for getting out of bed was from side lying, placing one or both hands in front of the body, leaning forward and pushing up to a sitting position.
Many activities are discouraged after cardiac surgery, such as weight carrying and exercises involving the pectoralis major. Few studies have been published evaluating which activities and exercises negatively affect the sternal incision [18–20]. The recommendation for how long after surgery the patients should avoid weight bearing and certain other activities, differs with a range of 7 to12 weeks. Likewise, how much weight patients are allowed to lift while the sternum is healing differs between 1 and 5 kg. It has been suggested that current activity guidelines for CABG patients are too restrictive ; however, considering that postoperative sternal instability is a serious complication with increased risk of mortality, the importance of correct instructions for sternal precautions is essential, especially in risk patients . More scientific knowledge of risk factors and risk behaviours for sternum instability is needed. This would provide further possibilities to individualize the postoperative recommendations to the patients.
All physiotherapists in the present study considered physiotherapy necessary after cardiac surgery, although one-third considered the physiotherapy treatment offered not optimal. The main reason mentioned was lack of time.
A national Swedish guideline for physiotherapy treatment for patients undergoing major surgery is currently under development, but was not available during the study period. In spite of this, the physiotherapy management given in the different departments, by different physiotherapists, was fairly similar. An explanation for this may be the yearly national meetings for physiotherapists in the cardiovascular field. This survey provides information that may be useful in research as well as development and implementation of clinical practice guidelines in physiotherapy. It is also very important to widen this knowledge and formulate internationally accepted guidelines for cardiac surgery patients.