Mariken Stegmann

Setting and study design  As in several other Western countries, Dutch GPs have a long relationship with their patients  and they provide 24/7 primary care service. The GP also acts as a care coordinator, being  responsible for managing all of their patients’ medical data and for liaising with specialists  from different services.  This study is designed according to the SPIRIT guidelines 18  and planned to be a two‐armed  parallel group pragmatic randomised trial. 19  We will study the effects of GPs eliciting  preferred treatment goals on patient self‐efcacy, among older patients with cancer in a  palliative setting, in comparison to care as usual. The primary endpoint is the perceived self‐ efcacy directly after consultation with an oncologist during which the choice for a specic  treatment has been made (Fig. 1).  Participants  Recruitment  Eligible patients will be recruited by oncologists from eight locations of ve hospitals (one  academic and four non‐academic) in the north of the Netherlands when a new treatment  decision is to be made (e.g. at the moment of diagnosis). The oncologist will then check the  inclusion and exclusion criteria, and if suitable, will provide interested patients with written  information about the study and ask for permission to send their contact details to the  researchers.  Inclusion and exclusion criteria  Participants will be required to be aged 70 years or older, have non‐curable cancer, and be in  a position requiring them to make a new treatment decision. We will exclude patients if they  have a life expectancy of   less than three months, have a severe psychiatric disorder; are  unable to ll out questionnaires; or have a hematologic malignancy (because the course of  these diseases differ from solid tumours 20 ). Patients formerly treated curatively will not be  excluded, neither will patients with other comorbidities.  Randomisation  After receiving informed consent, patients will be consecutively allocated to the intervention  or care as usual groups. The 1:1 allocation sequence will be produced by a web‐based  randomisation application using random permuted blocks with randomly varying block sizes  stratied by hospital. Allocations will then be kept in a locked le on a computer system,  inaccessible to the recruiting oncologists and the researchers.  Intervention  Care in intervention group  When patients are randomised to the intervention group, they will be asked to arrange an  appointment with their GP before the next scheduled consultation with their oncologist.  During this appointment with the GP the diagnosis of cancer in a palliative setting, and its  consequences will be discussed. Using a specifically designed “Outcome Prioritisation Tool”  (OPT), also called the “preference slider,” preferred treatment goals will then be  explored. 9,21  This OPT is a card with four visual analogue scales, each representing a generic  treatment goal (Fig. 2): life extension, preserving independence, reducing pain, and reducing  2 21 Protocol for the OPTion study