Comparing physician and nurse ECOG performance status ratings as predictors of clinical outcomes in cancer patients.
248 Background: The Eastern Cooperative Oncology Group Performance Status (ECOG) scale is commonly used in the clinical and research setting, and has been shown to correlate with cancer morbidity, mortality, and tolerance to chemotherapy. ECOG is frequently rated by physicians (MDs) and nurses (RNs), and MD-RN ECOG score agreement varies widely in the literature. It remains unclear whether MD and RN ECOG scores differ in their ability to predict clinical outcomes. Methods: As part of a quality initiative at Cedars-Sinai Cancer Center, oncologists and chemotherapy RNs independently scored ECOGs for a random sample of 309 patients with various solid malignancies, on one or more visits, for a total of 506 pairs of ECOG scores. MD/RN interrater agreement was evaluated using the Cohen's kappa coefficient. Logistic regression models were fit to evaluate the ability of RN and MD ECOG scores, as well as the RN-MD score difference, to predict the occurrence of chemotoxicity (CTCAE v4, grade≳3) and hospitalizations within 1 month from ECOG ratings as well as 6-month mortality/hospice referrals. Results: The agreement among 506 ECOG MD/RN pairs was 71% (Kappa = 0.485, p < 0.0001). RN ECOGs had a stronger odds ratio (OR) for 6-month mortality/hospice (OR = 3.55, CI 2.2-5.7) compared to MD ECOGs (OR = 2.99, CI 1.67-5.3). RN ECOG scores also significantly correlated with one-month chemotoxicity (OR = 1.39, CI 1.02-1.90), but MD ECOGs did not. Both MD and RN ECOG scores did not significantly correlate with 1-month hospitalizations. The magnitude of RN to MD ECOG score difference was also positively associated with 6-month mortality/hospice (OR = 3.42, CI 1.87-6.3), but not with 1-month hospitalization or chemotoxicity. Conclusions: Our findings suggest that RN-rated ECOGs may be stronger predictors of chemotoxicity and 6-month mortality/hospice compared to MD ECOGs. Furthermore, the magnitude of difference between ECOG MD and RN ratings was associated with increased mortality/hospice rates, specifically when the MD rating was “healthier” than that of the RN. As ECOG scores are frequently used for prognostication and to inform treatment decisions, ECOG scoring by RNs may result in additional clinical benefit.