Patients’ Understanding of Oncology Clinical Endpoints: Environmental Scan and Focus Groups

Abstract

Background

Understanding treatment options is important for patients with cancer and their caregivers. This may be difficult, however, because oncology treatments are often approved based on complex clinical endpoints. The study aimed to explore lay understanding of oncology clinical endpoints by assessing the definitions of clinical endpoints available online and gathering qualitative focus group data on cancer survivors’ and the general public's understanding of clinical endpoints.

Methods

We conducted an environmental scan to find Web sites accessible by a general audience that defined three clinical endpoints: overall survival, progression‐free survival, and response rate. Next, we conducted a series of eight focus groups across the U.S. with cancer survivors (n = 36) and general population adults (n = 36).

Results

We found several online resources defining each endpoint; however, many of the definitions we identified used technical language that may not be easily understood by patients and caregivers. Few focus group participants were familiar with the technical terms for these endpoints. When presented with the endpoint terms and definitions, participants had misconceptions about treatment efficacy. Specifically, they tended to expect that all endpoints were a variation on living longer.

Conclusion

The results point to the need for more patient‐friendly definitions of clinical endpoints developed with input from the general public and from patients with cancer.

Implications for Practice

As the number of oncology prescription drug approvals and the advertising of those drugs to consumers increase, it is timely and critical to understand how to discuss treatment benefits with patients. Patient‐friendly definitions of common clinical endpoints, such as overall survival and progression‐free survival, would help health care providers describe treatment benefits to patients. This research provides evidence regarding patients’ understanding of these endpoints and suggests definitions for additional research. This represents a first step in creating evidence‐based patient‐friendly language to describe clinical endpoints.

Short abstract

Understanding treatment options can be difficult for patients with cancer and for their caregivers. This article explores lay understanding of oncology clinical endpoints, assessing definitions and focus group data on cancer survivors’ and the general public's understanding of clinical endpoints.

Introduction

Many patients with cancer want information on treatment options as part of the decision‐making process [1, 2]. However, communicating treatment options can be difficult, partially because treatments are often approved based on complex endpoints, such as progression‐free survival and response rate [3]. A recent literature review found few studies on the topic of communicating clinical endpoints to patients [4]. The little evidence available suggests that when patients receive information on these endpoints it can lead to confusion about the treatment's purpose, with some patients mistakenly believing a treatment will lead to longer life or a cure. For instance, Fallowfield and colleagues interviewed 90 oncology patients who were prescribed drugs that had demonstrated only progression‐free survival or modest overall survival benefits [5]. When asked about the aims of their treatment, most patients (92%) correctly believed that an aim of treatment was to slow or stop their cancer, but 50% also believed that this meant living longer. This underscores the need for patient‐friendly definitions of clinical endpoints. Indeed, a recent article specifically called for more research on patient‐friendly definitions of clinical endpoints to help patients understand oncology treatment options [6].

This study explored lay understanding of oncology clinical endpoint information. We focused on three specific clinical endpoints: overall survival, progression‐free survival, and objective response rate. Overall survival is the most reliable direct measure of effectiveness in oncology clinical trials. Surrogate endpoints like progression‐free survival and objective response rate are often measured before overall survival can be measured or when measuring overall survival is complicated by other factors [3, 7]. We chose to focus on these endpoints because they are commonly used in clinical trials, prescription drug labeling, and direct‐to‐consumer prescription drug advertising [7, 8, 9].

Materials and Methods

First, we conducted an environmental scan to assess the definitions of these endpoints that patients and the general public may encounter when searching for cancer information online. Second, we conducted exploratory qualitative research with general population adults and cancer survivors in a series of eight in‐person focus groups to explore reactions to and understanding of these clinical endpoints. Focus groups are ideal for generating in‐depth discussion about individuals’ perceptions and behaviors, and they enable rapid data collection [10]. This type of in‐depth information is difficult to capture in quantitative studies or large‐scale surveys, especially if response categories are not well defined or little is known about the topic. We used the definitions identified in the environmental scan to draft patient‐friendly definitions for use in the focus groups.

Environmental Scan

In February and March of 2019 we conducted an environmental scan to find Web sites accessible by a general audience that defined three clinical endpoints: overall survival, progression‐free survival, and response rate. Following previously established methods [11, 12], we created a list of search terms (Table 1) and entered various combinations of these search terms into the Google search engine by (a) searching each endpoint on its own, (b) pairing each endpoint with terms that indicated a definition, and (c) pairing each endpoint with terms that indicated a nonmedical audience. This led to 15 independent searches for each endpoint. For each independent search, we reviewed the first 20 links. If we identified a Web site with an endpoint definition in the first set, we reviewed the next 10 links. If we identified another Web site, we reviewed the next 10 links. We repeated this process until no additional Web sites were identified.

Table 1.

Environmental scan search terms

Clinical endpoint terms	Definition terms	Audience terms
Overall survival Progression‐free survival Objective or overall response rate	Definition Define Describe Description Explanation Explain Means Meaning What is	Consumer Patient General public General population Lay

Focus Groups

We conducted a series of eight 1‐hour focus groups in focus group facilities in four U.S. cities that were selected to achieve geographic diversity (Minneapolis, MN; Raleigh, NC; Philadelphia, PA; and Walnut Creek, CA). In each city, we conducted one nine‐person focus group with a general population of adults and one nine‐person focus group with adults diagnosed with cancer who had completed treatment (n = 72).

Eligibility Criteria

All participants had to be 18 years of age or older, English‐speaking, with no recent (past 3 months) focus group or interview participation. Participants had to report never working for the Department of Health and Human Services, a pharmaceutical company, a marketing or market research company, a health care company, or in the health care field. Participants had to be comfortable talking to their doctor [13], as measured by the following item (responses of 4 or 5 were eligible): “On a scale of 1 to 5, with 1 being ‘not at all comfortable’ and 5 being ‘extremely comfortable,’ how comfortable are you with asking your doctor to explain something that you are confused about?” The eligibility criteria were designed to maximize the number of eligible participants who could speak sufficiently to each of the study's research questions.

We established specific eligibility requirements for the cancer survivor focus groups. Participants in these groups had to report a previous diagnosis with a solid tumor cancer (which included two forms of skin cancer, Kaposi sarcoma and melanoma) or hematologic cancer. They also had to report not currently receiving treatment for cancer. We recruited a mix of solid tumor and hematologic cancer survivors for each of the cancer survivor focus groups. Table 2 shows participant characteristics.

Table 2.

Focus group participant characteristics

Characteristic	General population (n = 36), n (%)	Cancer survivor population (n = 36), n (%)
Sex
Male	18 (50)	18 (50)
Female	18 (50)	18 (50)
Age, mean ± SD	45.9 ± 11.8	59.2 ± 11.4
Ethnic background
Hispanic	10 (28)	2 (6)
Non‐Hispanic	26 (72)	34 (94)
Race
African American	13 (36)	9 (25)
American Indian or Alaska Native	1 (3)	0
Asian	1 (3)	1 (3)
White	14 (39)	23 (64)
Other	7 (19)	3 (8)
Education
High school or less	16 (44)	11 (31)
Some college or technical school	8 (22)	9 (25)
College graduate	5 (14)	12 (33)
Postgraduate	7 (19)	4 (11)
Type of cancer a
Bladder	—	2 (5)
Breast	—	6 (16)
Chronic lymphocytic leukemia	—	2 (5)
Colon‐rectal cancer	—	2 (5)
Leukemia	—	2 (5)
Lymphoma	—	13 (35)
Melanoma	—	1 (3)
Ovarian	—	1 (3)
Pancreatic	—	1 (3)
Prostate	—	4 (11)
Thyroid	—	3 (8)

^aOne participant reported two cancers.

Recruitment and Screening

We collaborated with two market research firms to identify and recruit participants in each city. The firms identified potential participants in their respective cities via their contacts database and media advertisements. The firms then contacted potential participants by telephone and screened them for eligibility. If interested and eligible, individuals were scheduled for preselected days.

Data Collection

We developed a semistructured master moderator guide that contained a series of structured questions about key topics and a series of corresponding follow‐up questions and probes (supplemental online Appendix A). We organized the guide into several sections to address the key topics and started with a discussion of overall survival, followed by progression‐free survival and objective response rate. We included specific probes for cancer survivor focus groups to assess previous exposure to these topics throughout their treatment. Key topics included familiarity with terms, interpretations of terms, and reactions to definitions. We based the definitions on National Cancer Institute and American Society of Clinical Oncology definitions identified in the environmental scan, which we modified after reviewing definitions from other patient‐targeted Web sites and consulting with clinical advisors. Moderator guide development was driven by the research questions and guided by clinical advisors and qualitative methods experts.

Before each focus group, we provided participants with an informed consent form, explained the study procedures aloud, and answered any participant questions. A trained moderator used the structured guides to ask questions, probe for details, and lead participants in an open discussion about the endpoints. During the focus groups, participants also completed short exercises in a workbook to indicate points of clarity and confusion for each of the endpoint definitions. One note‐taker observed each focus group from a private observation room and documented major themes. Additional study staff observed the groups remotely via video‐streaming. We audio‐recorded focus groups (with consent). Participants received a cash incentive. The focus group procedures were reviewed by the relevant institutional review boards.

Data Analysis

We produced verbatim transcripts of each focus group and compiled the session notes. We then developed a coding scheme based on the study's objective, the moderator guide questions, and session notes. The first, second, and, for some codes, third levels of the coding scheme were fixed codes determined based on the moderator guide and study objectives; however, the third and fourth levels contained emergent codes that were created by the study team as they conducted additional analysis of the fixed codes to identify emerging themes across group and detect differences between the general population groups and the cancer survivor groups. We conducted the coding and analysis in phases reflective of grounded theory in qualitative analysis. This coding strategy allowed us to first organize participant responses into meaningful categories without predetermining specific “response options,” followed by additional coding to develop data‐driven emerging themes. This joint approach to coding—using a mix of concept‐driven and data‐driven codes—is common in qualitative research and is an effective way to balance a study's theoretical framework with the raw data collected [14, 15, 16].

Next, two study team members independently reviewed and coded one transcript in NVivo 11.0 qualitative analysis software to check for interrater reliability (IRR). Once IRRs of ≥0.70 were achieved, the same two study team members coded the remaining seven transcripts. These two team members examined participant responses, identified key emerging themes, and determined the degree of consensus or discordance with each view. Once coding was complete, we analyzed participant responses for trends across focus groups and, when applicable, within focus group segments (e.g., general population and cancer survivors). Participant workbook responses were analyzed to corroborate points of clarity and confusion identified in the verbal responses.

Results

Environmental Scan

Overall Survival

We identified 29 Web sites in the overall survival search. Eighteen Web sites included definitions related to the length of time (e.g., “The length of time from either the date of diagnosis or the start of treatment for a disease, such as cancer, that patients diagnosed with the disease are still alive”). Six of those 18 Web sites used technical language such as “randomization” (e.g., “Time elapsed between randomization and death from any cause”). Twelve Web sites included definitions for overall survival rate, which focuses on the proportion of people alive rather than the length of time (e.g., “An overall survival rate shows the percentage of people who are alive after a certain period of time after diagnosis of a disease”). Supplemental online Appendix B Table 1 presents these definitions.

Progression‐Free Survival

Of the 15 independent searches for progression‐free survival, one was stopped early: the search that combined “progression‐free survival” and “definition” was stopped after a review of 60 links; although the last 40 links revealed definitions, they were all in technical language. We identified 57 Web sites in the progression‐free survival search. Fifty‐five Web sites included definitions related to the length of time (e.g., “The length of time during and after the treatment of a disease, such as cancer, that a patient lives with the disease but it does not get worse”). Thirty‐three of those 55 Web sites used technical language (e.g., “The time from randomization until first evidence of objective tumor progression or death from any cause, with censoring of patients who are lost to follow‐up”). Three Web sites included definitions for progression‐free survival rate (e.g., “The proportion of people among those treated for a cancer whose disease will remain stable (without signs of progression) at a specified time after treatment”). Supplemental online Appendix B Table 2 presents these definitions.

Objective or Overall Response Rate

We identified 20 Web sites in the objective or overall response rate search. Four included definitions that mentioned tumor shrinkage (e.g., “The percentage of patients whose cancer shrinks or disappears after treatment”), and eight included definitions related to reducing tumor size or burden (e.g., “The proportion of patients with tumor size reduction of a predefined amount and for a minimum time period”). Eleven included definitions that mentioned complete and partial response (e.g., “The proportion of patients who have a partial or complete response to therapy”). Supplemental online Appendix B Table 3 presents these definitions.

Table 3.

Clinical endpoint definitions

Type of definition	Overall survival	Progression‐free survival		Objective response rate
	Solid tumor and hematology	Solid tumor–specific	Hematology‐inclusive	Solid tumor–specific	Hematology‐inclusive
Original publicly available definitions	The length of time from either the date of diagnosis or the start of treatment for a disease, such as cancer, that patients diagnosed with the disease are still alive. a	The length of time during and after treatment that the cancer does not grow or spread further. b	The length of time during and after the treatment of a disease, such as cancer, that a patient lives with the disease but it does not get worse. a	The percentage of patients whose cancer shrinks or disappears after treatment. a	N/A
Definitions used in focus groups	The length of time that patients are still alive after starting treatment.	The length of time after starting treatment that a patient lives with the disease but it does not grow or spread further.	The length of time after starting treatment that a patient lives with the disease but it does not get worse.	The percentage of patients whose tumor shrinks or disappears after starting treatment.	The percentage of patients with fewer detectable cancer cells or no detectable cancer cells in their body after treatment.
Suggested definitions for future research	The length of time that patients lived after starting treatment.	The length of time after starting treatment that the cancer does not grow or spread.	The length of time after starting treatment that the cancer does not get worse.	The percentage of patients whose tumor shrinks or disappears after starting treatment.	The percentage of patients with fewer or no detectable cancer cells after starting treatment.

Note. The findings suggest (a) providing details when possible (e.g., specific length of time, percentage, or treatment) and (b) potentially including additional information that distinguishes progression‐free survival and objective response rate from overall survival.

^aFrom the National Cancer Institute.

^bFrom the American Society of Clinical Oncology.

Focus Groups

Overall Survival

When presented with the term “overall survival,” about half of participants across all focus groups reported previous exposure to the term (n = 19 cancer survivors, n = 16 general population participants). Several participants expressed initial confusion about the term and said that they did not know what it meant. Among participants who provided a response, most associated the term with an increase in life expectancy as a result of taking a medication. Some participants noted that the drug may not act as a cure but will allow someone to live with the disease. Some participants reported associating overall survival with a specific rate, percentage, or specific survival time period that results from taking a medication. Less frequently, participants associated the term with a cure or solution to a cancer diagnosis.

We provided participants with a definition of overall survival (Table 3). Participants who reacted well to the definition thought it was clear and straightforward. However, most participants reported that certain words in the definition were confusing or that the definition could be improved by including more details, such as the length of time, type of treatment, and when the outcome was measured. Participants in most of the focus groups noted that the definition did not consider a patient's quality of life, which they felt was an important concept when discussing the life expectancy of patients with cancer.

Group 2 (cancer survivors): “The word ‘alive’ could be kind of confusing just in terms of what quality of life you have.”

Group 1 (general population): “Think it just speaks to the duration, speaks nothing to quality.”

One difference between the general population and cancer survivor groups was that more cancer survivor groups responded negatively to the “still alive” terminology than general population groups. Their main concern was that “still” sounded negative and was confusing or unnecessary. See Table 3 for revised definitions.

Group 6 (cancer survivors): “It's a little harsh. [How so?] Well, the lengths of time that patients are still alive, it's a little harsh. When you're going through something you don't want to hear it like that unless you ask to hear it like that.”

Progression‐Free Survival

When presented with the term “progression‐free survival,” three cancer survivors reported previous exposure to the term (and all three heard the term from a source other than their doctor). Most participants associated the term with a slowing or stabilizing of the growth of a cancer. Several participants erroneously thought that it meant an increase in the chance of surviving, having a cure, or that they would have fewer side effects. A few other participants associated the term with a cancer that is in remission.

We first presented participants with a hematology‐related definition of progression‐free survival (Table 3). Participants in most focus groups felt that this definition conveyed the notion that patients were continuing to live with cancer; however, there was confusion around the phrase “does not get worse.” All focus groups had negative feedback on at least one element of the definition.

Participants who reacted well to the definition thought it was clear and straightforward, and made it clear that it was referring to patients continuing to live with cancer but whose condition was not worsening. However, based on comments made at different points during the discussion, it was often unclear whether participants understood that progression‐free survival was not synonymous with longer life. Indeed, some participants thought that this definition was more optimistic than the overall survival definition.

Group 5 (general population): “Whatever stage they're in, that they're going to be maintained at that stage.”

Group 6 (cancer survivors): “So you live with it and it won't get any worse, and you're not dead.”

Group 1 (general population): “I think this is pretty optimistic compared to [overall survival], if I was somebody with cancer, between overall survival and progression‐free survival, I'd want to be in this stage.”

Some participants wondered if “it” was referring to the cancer itself or the surrounding symptoms of the disease or side effects from the treatment. Participants also were concerned that the phrase “does not get worse” is too vague and confusing and could mean a variety of things.

Group 2 (cancer survivors): “It's a strange concept if you've got cancer to say, ‘Well, it's not getting worse.’”

We followed up with a second definition specific to solid tumors (Table 3). Although some participants had no preference, overall the participants largely preferred the second definition. Participants thought that it was more specific and less vague than the original definition.

We asked participants if the term “cure” comes to mind if a drug is shown to improve progression‐free survival. Overall, participants did not associate an improvement in progression‐free survival with a cure; however, a few participants suggested that improvement in progression‐free survival could be a sign that a cure is close. Across focus groups we saw mixed responses, with some participants noting that improvements in progression‐free survival would lead to a patient feeling better, whereas others noted that it does not mean that a patient would feel better or necessarily live longer.

When comparing the general population and cancer survivor focus groups, we found the latter were more likely to ask for clarification about the type of treatment referenced in the definition (i.e., whether “treatment” referred to chemotherapy, surgery, radiation, a regimen of cancer medications, or their most recent cancer treatment).

Objective Response Rate

When presented with the term “objective response rate,” one cancer survivor reported previous exposure to the term. Participants shared thoughts that objective response rate refers to how soon the drug works, how effective or responsive the drug is, and how an individual responds to treatment. Some participants were confused by the word “objective” (vs. “subjective”). The word “rate” was also confusing, as some participants thought “rate” had to do with the amount of time to respond to treatment.

We first presented participants with a definition of objective response rate specific to solid tumors (Table 3). Many participants commented that there was a big difference between the terms “shrink” and “disappear” and found it odd to include both in the definition. Some participants noted that they would prefer a drug that made a tumor disappear (vs. shrink) and other commented on how the word “shrink” was vague. Some participants noted that they would want to know by how much the tumor shrank if it did not disappear.

Group 4 (cancer survivors): “Shrinks seems to be more subjective than objective. Disappearing I don't think is, I mean, I think I could see that as black and white, you either have it or you don't have it.”

A few participants brought up questions about side effects and wondered how the drug might affect quality of life for patients.

Group 3 (general population): “If I were to add anything I would say, well, did the tumor shrink or disappear without negatively affecting the surrounding tissues or without negatively affecting, you know, the quality of life. … So again, to clarify that it's not negatively affecting the rest of you while it's doing that might be helpful.”

We followed up with a second definition more inclusive of hematologic cancers (Table 3). Participants had similar reactions to this definition and commented that the definitions were similar. A few participants did not like “fewer” and preferred a more specific value, and some asked how many fewer cancer cells would make a difference. Some participants expressed confusion over the word “detectable” because it made them wonder if there were cancer cells in the body that are undetectable.

When comparing the general population and cancer survivor focus groups, we found that, unprompted, a few participants in cancer survivor focus groups commented that they thought that the definitions implied a cure. When we specifically asked if improvement in objective response rate brings “cure” or “live longer” to mind, participants in two focus groups (one cancer survivor group and one general population group) tended to all agree that it did, and one general population focus group all agreed that it did not. Participants in three other focus groups (two general population and one cancer survivor population) had mixed responses, as some participants did not associate the definition with a cure, but slightly more did. Among the participants who thought of a cure, they did so based on the word “disappears” or “not detectable.”

Group 8 (cancer survivors): “If it disappears, I would think a cure.”

Group 4 (cancer survivors): “But to begin with I just think the cure is more of a definite, you know, it's gone, but when you get shrinks in there, you know, you just have a little bit more waiting, you can't put anything as a definite.”

Discussion

We conducted an environmental scan to identify Web sites accessible by a general audience that define three clinical endpoints: overall survival, progression‐free survival, and objective or overall response rate. We found that endpoint terms and definitions were readily available online. Even though health care professionals may not be discussing these endpoints with patients [4], previous research has found that patients with cancer actively seek information online [17]. However, many people are still frustrated by their online health searches [18]. In line with this finding, many of the definitions we identified used technical language that may not be easily understood.

The environmental scan provides information on the type and amount of information available online, but it does not tell us how individuals react to or understand this information. Therefore, we conducted a series of focus groups to determine how individuals from the general population and cancer survivors understand and react to the three clinical endpoints of interest. Unsurprisingly, few focus group participants were familiar with the technical terms for these endpoints. When presented with the endpoint terms and definitions, participants had misconceptions about treatment efficacy. Many either said or intimated that the terms indicated a cure or longer life. Several participants were skeptical about whether these terms encompassed quality of life.

Overall, cancer survivors were more likely to be familiar with the terms, and they were more likely to ask for more details, like the specific treatments alluded to in the definitions. But, surprisingly, we did not find many differences between the general population and cancer survivor groups in their reaction to and understanding of the terms.

Study Limitations

The environmental scan provides a snapshot of the information available online at the time it was conducted; this information may change over time. We also limited our search to three clinical endpoints; however, there are several other endpoints that may be relevant to patients with cancer, such as disease‐free survival and patient‐reported outcomes like quality of life.

The focus groups were conducted with a self‐selected sample, which limits the generalizability of the findings. In particular, individuals currently receiving treatment for cancer were not represented because a discussion of how improvements in clinical endpoints do not necessarily mean longer life may have been insensitive to these individuals. Nonetheless, a larger, more diverse sample may have uncovered additional reactions to these endpoints.

To avoid overburdening participants, we focused on five clinical endpoint definitions. Although the definitions we tested were similar to those we identified in the environmental scan, we did not test verbatim publicly available definitions. Future research could test multiple publicly available definitions to determine their relative clarity.

It is possible that the order in which we presented the endpoints affect some participants’ responses and prompted them to ask particular questions. Finally, participant responses could have been influenced by group discussions and dynamics. However, in this study, there was no evidence of a dominant individual influencing any group, and participant behavior indicated an open discussion climate.

Discussions between health care providers and patients about the efficacy of a specific treatment are part of a broader conversation about treatment goals. As researchers work toward creating patient‐friendly definitions of clinical endpoints, it is important to keep in mind that patients are often also confused about the overall goals of their treatment [19, 20, 21]. Future research could examine how discussing clinical endpoint definitions affects patients’ understanding of the curative or palliative intent of their overall treatment plan.

Clinical Implications

New oncology prescription drugs are entering the market, and they are often approved on the basis of surrogate endpoints such as progression‐free survival and objective response rate [3]. In addition, direct‐to‐consumer advertising of oncology prescription drugs is increasing in the U.S., and many of these ads include claims about clinical endpoints [4]. Examples include “Proven to offer a longer life,” “More than half of women saw their tumors shrink versus [another treatment],” and “significantly more effective at delaying disease progression versus [another treatment].” If these ads drive patients to talk to their physicians about these products, physicians may need to explain these clinical endpoints and what the drugs may or may not do for the patient.

These results suggest the need for more patient‐friendly definitions developed with input from the general public and from patients with cancer. When describing these endpoints, health care professionals and health communicators may need to be clear about what the endpoint means (e.g., more time before the tumor grows or spreads) and also what it does not mean (e.g., but not necessarily longer life or better quality of life) to help patients make informed treatment decisions.

Conclusion

Although there is a recognized need for patient‐friendly definitions of oncology clinical endpoints [6], little research has been done in this area [4]. Using definitions crafted from patient‐friendly resources identified online, we gathered focus group feedback from cancer survivors and members of the general public on their understanding of these endpoints. This research represents a first step in creating patient‐friendly language to describe clinical endpoints. Further qualitative research with different patient populations and different clinical endpoints, as well as quantitative studies examining the effect of this language in the context of health communications, will complement these findings and further the goal of developing language to help patients understand treatment options.

Author Contributions

Conception/design: Helen W. Sullivan, Amie O'Donoghue

Provision of study material or patients: Helen W. Sullivan, Kate Ferriola‐Bruckenstein, Janice P. Tzeng, Vanessa Boudewyns

Collection and/or assembly of data: Helen W. Sullivan, Kate Ferriola‐Bruckenstein, Janice P. Tzeng, Vanessa Boudewyns

Data analysis and interpretation: Helen W. Sullivan, Amie O'Donoghue, Kate Ferriola‐Bruckenstein, Janice P. Tzeng, Vanessa Boudewyns

Manuscript writing: Helen W. Sullivan, Amie O'Donoghue, Kate Ferriola‐Bruckenstein, Janice P. Tzeng, Vanessa Boudewyns

Final approval of manuscript: Helen W. Sullivan, Amie O'Donoghue, Kate Ferriola‐Bruckenstein, Janice P. Tzeng, Vanessa Boudewyns

Disclosures

The authors indicated no financial relationships.

Supporting information

See http://www.TheOncologist.com for supplemental material available online.

Appendix S1: Supplementary Information