Who really knows the performance status: The physician or the patient?

An integral part of patient-centered care in oncology is shared decision making regarding treatment options. To have these conversations, clinicians must understand a patient’s overall state of health and his or her values. In practice, one of the most important ways in which this is captured is a patient’s performance status (PS). PS is a ubiquitous and critical metric for prognosticating and determining both medical and surgical therapy options for individuals with cancer. In this editorial, we discuss the historical and current utility of PS assessment and highlight ways to better understand and more accurately characterize PS in the future. The most commonly used measure of patient PS to date is the Eastern Cooperative Oncology Group (ECOG) criteria. First developed in 1974, the ECOG criteria were originally developed as a set of standardized toxicity and treatment response criteria (each on a 0-5 scale) for use in clinical trials of cancer therapies. Ironically, it is a third metric, a brief afterthought in the original work, that has stood the test of time and gained tremendous importance: “other patient status criteria,” that is, the PS. This metric, mirroring the Karnofsky Performance Scale, is an evaluation of a patient’s physical abilities, with “fully active” scoring 0, “completely disabled” scoring 4, and “dead” scoring 5. Because even 1 point on the scale has profound implications for prognosis, this number can often be an important deciding factor among different forms of treatment. For example, many cancer surgeons would hesitate to operate on a patient with a PS of 2. In a patient with a PS of 3, many medical oncologists may pause before giving chemotherapy and instead consider palliative care. The ECOG PS is typically scored by providers during clinic visits. Emerging research calls this practice into question by demonstrating that provider ratings of PS may be discordant from patient ratings. Bergerot et al took a critical look at provider and patient reports of ECOG PS in patients with metastatic genitourinary cancer. They found that providers were more likely than patients to give an ECOG score of 0 (92.4% vs 64.1%; P = .001). Interestingly, they demonstrated that a worse patient-reported ECOG PS was associated with worse physical and psychosocial well-being, whereas physician measures accounted only for the former. Thus, clinicians use the more readily visible and accessible physical status of a person, whereas patients also consider their mental, social, and emotional well-being when reporting their PS. In other words, patients are more holistic when they consider their own health status. As providers caring for patients with cancer, we have a critical opportunity to do a better job in examining the PS of patients holistically. Perhaps this begins with an in-depth examination of the most frequently used tool in our arsenal: the ECOG PS. It is undoubtedly a practical metric because of its brevity, universal availability, and validated prognostic power. However, the 5-point scale developed almost half a century ago may not capture the nuances of the patient experience in today’s oncology treatment landscape. The scale’s succinctness and lack of gradation run the risk of introducing heuristics that lead to inaccurate or incomplete judgments by clinicians. When looking to gather the most complete picture of a patient with cancer, we must consider who should be making PS assessments. We know that patient-reported outcomes (PROs) play an integral role in cancer care. These can include canceror symptom-specific rating scales, such as the UCLA Integrated Staging System for renal cell carcinoma or the International Prostate Symptom Score for prostate disease. Perhaps, as suggested by Bergerot et al, we should more frequently use the ECOG PS as a patient-reported metric in conjunction with other PROs that may capture a more comprehensive picture of PS. A multitude of factors beyond physical health can affect a cancer patient’s overall health and disease trajectory and must be considered when PS is being assessed. Prior work has shown that mental health factors can contribute to differences in provider and patient ratings of PS. Bergerot et al used the Patient-Reported Outcomes Measurement