Patient-reported and clinician-rated performance status and general health among women with gynecologic cancers on palliative chemotherapy.

Abstract

e24128 Background: Oncologists’ perceptions of patients’ performance status (PS) and general health are key determinants of eligibility for chemotherapy and clinical trials. We sought to examine correlations between patient-reported and clinician-rated Eastern Clinical Oncology Group PS (ECOG-PS) and general health status (EQ5D) in women with incurable recurrent gynecologic cancers, as well as step counts from accelerometry as a proxy for objective PS. Methods: We used baseline data from a randomized trial evaluating a mobile health intervention that collects patient-reported outcomes and passive smartphone data and provides tailored symptom management advice. Half of participants also received a wearable accelerometer. We calculated Pearson correlations to determine agreement between ECOG-PS and EQ-5D assessments in patient-clinician dyads. We also calculated the median (IQR) step counts among a subset of women with ≥7 days of ≥8 hours wear time. Results: Data from 78 patient-clinician dyads demonstrated a moderate correlation between patient-reported and clinician-rated ECOG-PS (r = .47; P< .000). Among dyads, 38 (49%) had perfect agreement; 36 (46%) had higher patient-reported ECOG-PS, indicating more patient-reported functional limitations, compared with clinician reports; and 4 (5%) had lower patient-reported ECOG-PS. Comparisons between patient-clinician ratings of EQ-5D subscales indicated moderate agreement for mobility (r = 0.44; P< .001), pain/discomfort (r = 0.45; P< .001), and usual activities (r = 0.36; P= .002); and weak or no agreement for anxiety/depression (r = 0.31; P= .007) and self-care (r = 0.02; P> .05). Participants’ daily step counts varied widely (median = 5,511 steps, IQR = 3,784, range 1,736-13,961) and were not associated with patients’ or clinicians’ ECOG-PS estimates (r = -.12; P= 0.44 and r = -0.03; P= 0.84) or patient-reported mobility (r = -0.10; P =0.51). Conclusions: Among women with incurable gynecologic cancers, patients’ and clinicians’ perceptions of patients’ ECOG-PS were moderately correlated, although 46% of clinician ratings were more optimistic than patients’ ratings. There was also moderate agreement between dyads for mobility and pain/discomfort. Patients’ step counts were not correlated with ECOG-PS or patient-reported mobility. Future studies should examine these measures longitudinally to determine which is the most sensitive predictor of clinical outcomes. Clinical trial information: NCT03022032 .