Introduction Discrepancies between physician-assessed and patient-assessed performance status, have been associated with an increased risk of mortality. Quantification of activity through actigraphy may be a better, objective method for prognostication. The actigraphy-derived dichotomy index (I<O) has been reported to be of prognostic value in the metastatic colorectal cancer and advanced lung cancer populations. However, the association between the I<O and ECOG-PS has not been fully explored.

Objectives To assess the relationship between physician-assessed ECOG-PS scores and the I<O and other actigraphy-derived activity parameters such as daytime activity and mean 24 hour activity.

Methods 50 adult outpatients with advanced cancer and an estimated prognosis of less than a year were recruited as part of a feasibility study. Patients and the palliative care physician independently assessed the patient‘s ECOG-PS both at baseline and after 7 days. Participants were instructed to wear an Actiwatch Spectrum Plus for seven consecutive 24 hour periods on their non-dominant arm, and to concurrently complete a sleep diary.

Results On Day 8, there was moderate agreement between the palliative care physician and individual patient‘s assessment of their ECOG-PS, with a Kendall’s correlation of 0.70 (p<0.001). A moderate negative correlation was observed between physician-assessed ECOG-PS and the dichotomy index (I<O) (r=−0.55; p=0.0003). There was no correlation between physician-assessed ECOG-PS and mean daytime activity (r=−0.29; p=0.073) or mean 24 hour activity level (r=−0.2; p=0.218).

Conclusions Physician-assessed ECOG-PS and patient-assessed ECOG-PS scores are moderately correlated. A poor performance status is significantly associated with a measure of day-night difference in activity, but not with absolute activity measures.