In the course of their disease, most patients with Parkinson's Disease (PD) face mounting mobility deficits, including difficulties with walking, balance, posture and transfers. This frequently leads to (fear of) falls, injuries, loss of independence, and inactivity which causes social isolation and increases the risk of osteoporosis or cardiovascular disease. These mobility deficits are difficult to treat with drugs and neurosurgery. However, physiotherapy is deemed effective in improving mobility deficits in PD. Physiotherapy is widely prescribed for this purpose in the Netherlands. Yet, the efficiency of current "usual care" physiotherapy can be questioned, for two reasons. First, the referral process seems inadequate because patients are mainly referred by neurologists who often lack insight into the (im-)possibilities of physiotherapy for PD. Consequently, patients with a real need for physiotherapy are not always referred (undertreatment), whereas others without a real need are (overtreatment). Furthermore, most therapists treating PD patients are not specifically trained in treating these patients. This is not surprising because average therapists rarely treat more than two patients per year in their practice. Therefore, patients who are being referred probably receive suboptimal treatment.

The objective of this study is to evaluate whether the efficiency of physiotherapeutic care for patients with Parkinson's disease can be improved, at a reduced cost, by targeting two key elements of the current care system: a) inadequate referral by neurologists; b) suboptimal treatment by physiotherapists. We expect that optimal referral combined with expert treatment will increase the efficiency, as reflected by increased health benefits for patients at equal or reduced costs'.

Design In a Cluster Randomised Trial, 16 clusters will be randomly allocated to either network care (8 clusters with an altered organisation of physiotherapeutic care) or usual care (8 clusters with unchanged organisation of physiotherapeutic care). Clusters are formed by all PD patients living in the communities connected to participating regional hospitals in the 16 clusters.

The health care intervention in the experimental group has two elements: (a) an improved quality of referrals by neurologists; and (b) an improved quality of interventions by physiotherapists. Brief description Network Care: In each of the Network Care clusters, 5 to 7 motivated therapists are selected to enroll in a regional ParkNet and consequently trained. Training is focused at correct use of the evidence-based guidelines for physiotherapy in PD (Keus et al, 2006). This training consists of a 5-day competence-oriented course, web-based continues education supported by seminars, and use of a PD specific electronic patient record. Neurologists are informed about indications for referral to physiotherapy. Improved communication between neurologist and ParkNet therapists is initiated and supported.

Following implementation of the health care change, PD patients attending the neurological outpatient clinics of the individual hospitals within the clusters will be asked to participate. During a period of 6 months, PD patients will enrol in the study. Enrollees will be followed for 6 months to measure the use and quality of physiotherapy, patient health benefit and satisfaction, and costs.

Inclusion Criteria:

• Patients with idiopathic PD, diagnosed according to the Brain Bank criteria of the UK Parkinson's Disease Society
• Living independently in the community
• Able to complete the trial questionnaires.

Exclusion Criteria:

• Atypical parkinsonian syndromes
• Hoehn & Yahr stage 5
• Severe cognitive impairment
• Presence of major psychiatric disorders
• Severe co-morbidity (e.g. cancer) that interferes with daily functioning.