The aim of this project is to explore if monitoring physical activity including feedback can have any effect on physical activity level, function, work ability, health related quality of life or work productivity among individuals with hip and/or knee osteoarthritis. This will be performed by evaluating the use of mobile technology and activity monitoring to support physical activity in individuals with OA in a cluster randomized controlled trial. The investigators research questions are:

• Can an intervention, comprised of the above mentioned technology, have any effect on physical activity level, function, work ability, quality of life or work productivity among individuals with OA?

Methods Design and setting The project will be performed as a cluster randomized controlled intervention study within primary health care in the region of Skåne and Halland.

Sample Patients who are referred to the patient education program for OA in primary health care will be invited to participate in the project. Participants will also be recruited using advertisement on social media. Participants recruited from social media will attend the patient education program for OA arranged by this project.

There will be one control group and one intervention group, see details below:

A. Patient education program and physical activity monitoring. C. Patient education program (control).

Each group within the patient education program for OA will be randomized to either group A or group C. Patients that fulfil the inclusion criteria will then be asked to participate. Digital (BankID or e-legitimation) informed consent will be obtained at the project's website. Recruitment of participants in intervention group A and control group C starts in autumn 2017 and continues through spring 2019 or when sufficient numbers of participants have been included.

Interventions Activity monitoring will be performed by giving each participant in group (A) a wearable sensor (Fitbit Flex 2), where the daily physical activity level will be recorded. The monitoring will start at the patient education program within BOA. All activity monitors will be programmed with a daily goal of 7 000 steps. The participants will have the opportunity to follow their physical activity through the fitbit app were they also will receive standardised feedback. The intervention will be performed during three months, with follow-ups online after 3, 6 and 12 months.

Measures Patient-reported outcomes

Participants will answer the following questionnaires at baseline, after the intervention (3 months) and 6 and 12 months after baseline. Measurements will include:

• Self-rated work ability (Work ability Index, Ilmarinen,2007)
• Self-rated work productivity (WPAI-OA, Legget et.al.,2016)
• Self-rated health-related quality of life (EQ-5D) (http://www.euroqol.org/)
• Self-rated function in relation to hip and knee arthrosis (HOOS, Nilsdotter 2003, KOOS, Paradowski 2006).
• Self-rated physical activity (IPAQ, Craig et.al.2003).
• Questions about the work environment and physical activity.

Statistics The investigators will recruit approximately 80 participants/group. With this sample size and 80 % power, effect sizes of about 0.45 SD (significance level 5%) can be identified. In terms of WAI, this corresponds to SD 6, approximately 2,7 points increase/higher score in the intervention group, compared to the control group. Compliance with the intervention will comprise of the activity monitoring device used on at least 50% of the work days. With relatively large group sizes and most data expected to a follow an approximately normal distribution, group differences in the development over time with respect to the different outcome variables will be analyzed by ANOVA models (repeated measures or cross sectional comparison of mean difference scores). Some outcome variables, may likely be skewed and hence appropriate transformations, or non-parametric models, will be applied. Final and exact analytical strategies will be decided once data is available for exploration of final group sizes and distributions/symmetry of numeric outcome variable data.

Inclusion Criteria:

• Working at least 50 %

Exclusion Criteria:

• Not being able to do physical activity, non-Swedish speaking

Masking is performed through sealed envelopes. The investigator has no knowledge of which patients who is randomized to intervention or control