Should patient-rated performance status affect treatment decisions in advanced lung cancer?

INTRODUCTION: The Eastern Cooperative Oncology Group (ECOG) score is a well known prognostic factor and almost always used to determine eligibility for clinical trials. The patient-rated performance status score (Pt-PS), section of the patient generated subjective global assessment scale, has identical criteria to the physician-rated ECOG scale (MD-PS). We compared the Pt-PS with MD-PS in patients with advanced non-small cell lung cancer and compared the effect of each rating on eligibility for a hypothetical clinical trial.
METHODS: Consecutive patients with newly diagnosed advanced non-small cell lung cancer completed a patient generated subjective global assessment self-rated questionnaire, which was then correlated (kappa statistic) with the ECOG PS recorded at the same time. Patients were treated with standard chemotherapy. Survival was determined using Kaplan-Meier statistics.
RESULTS: One hundred nine patients (M:F-54:55) were recruited. Pt-PS differed from MD-PS in 59 (54%) instances (p = 0.0001). When scores were not congruent, 41/59 (69%) patients evaluated themselves as having a worse PS than the physician's rating. Pt-PS was 0 to 1 in 60 (55%) patients whereas MD-PS was 0 to 1 in 78 (72%) patients. The functional status irrespective of evaluator was predictive of survival (p = 0.001 for MD-PS and p = 0.001 for Pt-PS). However, the median survival in those with MD-PS >/=2 was 3.3 (CI; 1.7-4.9) months whereas individuals with Pt-PS >/=2 had a median survival of 6.2 (CI; 5.4-6.9) months.
CONCLUSIONS: Pt-PS and MD-PS were not congruent in over half of the cases, with Pt-PS scores usually poorer. Almost half the patients would have excluded themselves from a hypothetical clinical trial (Pt-PS >/=2). This requires prospective evaluation.