Longitudinal patient-reported performance status assessment in the cancer clinic is feasible and prognostic.

PURPOSE: Performance status is prognostic in oncology and palliative care settings. Traditionally clinician rated, it is often inconsistently collected, recorded, and measured, thereby limiting its utility. Patient-reported strategies are increasingly used for routine symptom and quality of life assessment in the clinic, and may be useful for tracking performance status. METHODS: Tablet personal computers were used to collect patient-reported reviews of systems via the Patient Care Monitor (PCM) v2.0 for 86 patients with advanced lung cancer. Relevant subscales included the PCM Impaired Performance and Impaired Ambulation scales. Trained nurse clinicians measured performance status using traditional Karnofsky and Eastern Cooperative Oncology Group (ECOG) instruments. Correlation coefficients were used to compare performance status scales, and survival analysis was performed by Cox proportional hazards modeling. RESULTS: All four performance status scales demonstrated excellent internal consistency and convergent validity. Initial KPS and ECOG scores were statistically correlated with survival, whereas PCM scores showed a nonsignificant trend in this direction. Change in PCM Impaired Performance over time was statistically correlated with survival (hazard ratio = 1.62, P = .046), whereas the other three performance status measures were not statistically prognostic. CONCLUSION: Patient-reported performance status as measured by PCM v2.0 is at least as reliable as KPS or ECOG. The enhanced resolution provided by this patient-reported method allows for the detection of clinically meaningful changes in trajectory over time, potentially serving as an early-warning system to trigger clinical interventions. Further study is needed to test these findings on a larger scale.

Duke Scholars

Author Thomas William LeBlanc Medicine, Hematologic Malignancies and Cellular Therapy

Author Rebecca A Shelby Psychiatry & Behavioral Sciences, Behavioral Medicine & Neur ...

Author Gregory P. Samsa Biostatistics & Bioinformatics, Division of Biostatistics