2012 EDUCATIONAL BOOK - American Society of Clinical Oncology

Capturing the Patient Perspective:
Patient-Reported Outcomes as Clinical
Trial Endpoints
Deborah Watkins Bruner, RN, PhD, Benjamin Movsas, MD, and Ethan Basch, MD
Overview: Just as clinical trial design and rigor have evolved
with improvements in methods and processes, so too have
methods for capturing patient data in clinical trials. Substantial
evidence suggests that standard physician reporting of
symptoms for which we lack objective diagnostics (e.g., pain)
is often discordant with patient self-report. Current reporting
using the National Cancer Institute Common Terminology
Criteria for Adverse Events (CTC[AE]) for symptom capture
relies on a filtering system, from patient to physician to
medical record to medical record abstraction to data entry,
with each step requiring interpretation and the possibility of
error. In contrast, patient-reported outcomes (PROs) eliminate
the filter and rely on direct report. Furthermore, the lack of
validation and training in use of the CTC(AE) creates an
inadequate data capture system. Inadequacies might be observed
as underreporting or overreporting symptom preva-
THE TITLE of this chapter, “Capturing the Patient
Perspective: Patient-Reported Outcomes as Clinical
Trial Endpoints,” implies that the patient perspective is
separate from other clinical trial endpoints. All endpoints
involve patient data. Why then the emphasis on the patient
perspective? Currently, all cancer clinical trials use the
National Cancer Institute (NCI) Common Terminology Criteria
for Adverse Events (CTC[AE]), a descriptive system
that is used for adverse event reporting. The CTC(AE) uses
a grading (severity) scale for each adverse event term. 1
Grading is performed by either the physician or research
assistants, who abstract information from the patient’s
medical records. Either method of grading requires interpreting
or filtering the patient experience. In contrast, “A
patient-reported outcome (PRO) is a measurement of any
aspect of a patient’s health status that comes directly from
the patient (i.e., without interpretation of the patient’s
responses by a physician or anyone else).” 2
The Pros and Cons for Use or Nonuse of PROs in
Clinical Trials
What do we gain over standard CTC(AE) reporting by
including the patient perspective in clinical trials? Table 1
lists the complementary benefits of the CTC(AE) and PROs.
However, given limited resources, what is the downside of
experienced physicians and research associates grading a
patient’s symptoms? The downside is they might not be
correctly grading symptoms and/or they might be missing
key information; either can occur for a host of reasons.
Lack of Reliability and Validity of the Current
Grading System
The CTC(AE) was not designed as a physician checklist or
patient interview, and there are a variety of ways in which
the grading is performed and by whom it is performed,
which are described elsewhere. 3 This variation in grading
leads to questions about training to use the grading system,
and the answer is there is no training. This lack of training
lence and severity compared with patient self-report.
Inaccuracies in symptom reporting can lead to missing important
prognostic information, lack of understanding of patient
adherence with therapies, and lack of information for patient
decision making. They can also lead to opportunities lost in
terms of labeling claims and comparative effectiveness analyses.
New developments in patient-reported outcome (PRO)
reporting, including the PRO-CTC(AE) and models for incorporation
of PROs in clinical trials, might facilitate routine
PRO reporting complementary to CTC(AE) in clinical trials. In
addition, the cadre of validated PRO instruments already in
existence allows for more in-depth, hypothesis-driven evaluations.
For standard toxicity reporting, the time has come for
mandatory routine PRO symptom reporting complementary to
the CTC(AE).
is evident in the rare instances when the CTC(AE) has been
tested for interrater reliability, which is modest at best.
Kabaet al 4 evaluated the reliability of CTC version 2.0 with
five experienced research associates independently reviewing
17 different medical records and grading toxicities.
Agreement among raters was lowest for symptoms that are
difficult to directly observe and highest for symptoms that
have an associated laboratory test, for example, agreement
for nausea was 0.47 (0.23–0.71, 95% CI) and agreement for
febrile neutropenia was 0.88 (0.73–1, 95% CI). 4 Atkinson
and colleagues 5 assessed interrater reliability of multiple
CTC(AE) symptoms. In a study at an NCI-designated comprehensive
cancer center, patients receiving active chemotherapy
routinely had toxicity assessments performed by
two clinicians. The ratings were performed independently,
without access to the other physicians’ reports, on a medical
record form that included the name of each symptom, a
checkbox to note the grade of the toxicity, and a key with
definitions of each CTC(AE) grade for each symptom. Agreement
among raters was low for most symptoms (e.g. diarrhea
[0.58], constipation [0.5], nausea [0.52], and vomiting
[0.46]). In addition to poor interrater reliability, no assessments
of validity of the CTC(AE) have been published,
meaning it is unknown whether grade 1 is clinically different
from grade 2, etc. By contrast, the use of PROs in clinical
trials is held to stringent and rigorous development and
psychometric validation as described in detail in the U.S.
Food and Drug Administration (FDA) guidance for use of
PROs. 2
From the Nell Hodgson Woodruff School of Nursing, Winship Cancer Institute, Emory
University, Atlanta, GA; Radiation Oncology Department, Henry Ford Health System,
Detroit, MI; Memorial Sloan-Kettering Cancer Center, New York, NY.
Authors’ disclosures of potential conflicts of interest are found at the end of this article.
Address reprint requests to Deborah Watkins Bruner, RN, PhD, Nell Hodgson Woodruff
School of Nursing, Winship Cancer Institute, Emory University, 1520 Clifton Rd., Room
232, Atlanta, GA 30322-4201; email: dwbrune@emory.edu.
© 2012 by American Society of Clinical Oncology.
1092-9118/10/1-10
139