Movement sensors have been touted as providing the next generation of health care through objective measurements that will replace "subjective" health assessments and remove the fallible memory of patients from health decision making. However, assessing the level of disability and assessing the efficacy of treatment are iterative and constructive acts reliant on an alignment of shared perceptions between clinician and patient. The challenge then is to utilize movement sensors as a resource for this co-interpretation as opposed to a replacement. The examples presented in this article are from fieldwork of Parkinson's patients' deep brain stimulation programming sessions. We employed a design research study that included observations of the noninstrumented movement assessment in 10 naturally occurring deep brain stimulation clinical programming sessions, design explorations of a sensor for evaluating upper limb movement, and the resulting uptake of the system in the assessment and co-interpretation practices by clinicians and patients through a deployment of this sensor in eight naturally occurring clinical assessments. Our findings highlight how sensors can provide much needed co-interpreted assessment of movement but sensors can also intrude on this process through clinician or sensor authority.