Contrary to expectations, total therapy duration was found to be overestimated more in individual therapy sessions than in circuit class therapy sessions.

There are two main implications of these findings. First, in terms of Lonafarnib price clinical practice, accurate quantification of therapy dose is important to allow for reflection on current practice and to measure changes in practice accurately. The National Stroke Foundation Clinical Guidelines for Stroke Management (2010) recommend that stroke survivors should be provided with as much opportunity as possible to engage in active task practice during the first six months after stroke. The results of this study showed that, on average, therapists overestimated active time by 28%, and underestimated rest time by 36%. This means, that in an hour-long therapy session, therapists believe their patients are active for 17 minutes more than they actually are. Conversely, patients are resting for 22 minutes longer than estimated. This finding is in line with other studies examining therapists’ accuracy of estimating therapy time (Bagley et al 2009). These findings suggest that when accurate data for therapy dose are required, such as for research or to monitor adherence

to clinical guidelines, more objective methods of measurement should be employed. For example, simple counting of repetitions of tasks or exercises has been used to describe therapy dosage in clinical trials (Birkenmeier et al 2010), and many stroke survivors in rehabilitation are able to accurately count repetitions BGB324 order of their own practice (Scrivener et al 2011). More detailed information about physical activity both SPTLC1 in therapy and across the day can be collected using activity monitors such as accelerometers. To date, the majority of studies using activity monitors have been conducted with ambulatory, community

dwelling stroke survivors (Alzahrani et al 2011, Manns and Baldwin, 2009, Rand et al 2009). Less is known about the accuracy of these monitors to detect activity in people early after stroke who may move very slowly, and activity monitors cannot provide information about the context and purpose of activity. Second, in light of these findings, one of the reasons therapy dosage studies have shown small effect sizes may be that many have relied on therapist estimations of therapy time. It is possible that if dose of therapy were more accurately quantified in these studies, a larger effect may have been detected. This is of course speculative, but serves to highlight the need for accurate quantification of therapy dosage in clinical trials. This study has several strengths: it involved multiple rehabilitation centres, examined both individual and circuit class therapy sessions, and involved clinicians with a range of experience.