Feasibility, Acceptability and Findings from a Pilot Randomized Controlled Intervention Study on the Impact of a Book Designed to Inform Patients about Cancer Clinical Trials

Abstract

Objective

To assess the feasibility, acceptability and changes in knowledge among cancer patients assigned to receive a 160 page book on experimental cancer therapies and clinical trials.

Methods

We enrolled 20 patients with cancer who had never participated in a clinical trial, and randomly assigned them to receive the book either during Week 1or Week 4 of the study. We collected baseline patient demographic and cancer related information as well as knowledge about cancer clinical trials at Week 0. Follow-up surveys were administered at Weeks 3 and 6 for both study groups. Comparisons were made within and between groups randomized to receive the book Early (at Week 1) to those who received it Later (at Week 4).

Results

One hundred percent of data were captured in both groups at baseline, which decreased to 77.8% by Week 6. The vast majority of participants found the book moderately or very useful (89% in the Early Group at Week 3 and 95.5% in the Late Group at Week 6). Within group pair-wise comparisons found significant difference between baseline and Week 6 in content-specific knowledge scores among participants in the Late Group (79% vs. 92.1%, p=0.01). Global knowledge scores increased significantly for variables reflecting knowledge that promotes decisions to participate in clinical trials.

Conclusions

Providing published reading material to patients with cancer is both feasible and acceptable. Offering information to patients about cancer clinical trials, using a book designed for patients with cancer may influence knowledge related to decision to participate in clinical trials.

Introduction

Advancements in cancer treatment depend primarily on patients' participation in clinical trials (1). Many studies have shown that adult participation in clinical trials is low among all patient groups including diverse race and ethnic groups, ages, and sex (1) with participation in Phase II and III trials ranging from 12-51% of those eligible (2-4). Barriers to participation include mistrust of researchers (5, 6), lack of knowledge about clinical trials and opportunities to participate (7-9), pre-existing treatment preferences, and characteristics of the specific protocols such as research designs (10). An important issue for many patients is whether or not the study offers active treatment in all study arms (11). One study reported that differences in attitudes towards decisional control and perceived benefits associated with the trial arms also influence patients' decisions to take part in cancer clinical trials (12).

As a result of research on barriers, intervention studies have implemented alternate trial designs, improved informed consent processes, addressed gaps in knowledge about clinical trials (10), used communication strategies that promote trust and confidence, which can be highly influential (13, 14), developed community outreach strategies, as well as direct recruitment in academic and community clinic settings (15-18). A meta-analysis by Mill, et al conducted in 2006 (10) suggested that additional research should focus on perspectives of patients who both accept and decline to take part in clinical trials and on interventions designed to include sensitivity to patient quality of life as well as practical and ethical issues (10). Another study by Meropol, et al, found that perceptions of outcomes related to participating in cancer clinical trials appear to differ between patients and physicians, highlighting the need for better communication that involves patient education (18).

We conducted a pilot study using a randomized wait-listed controlled design, to assess the feasibility, acceptability and changes in knowledge and attitudes among cancer patients assigned to receive a book entitled, “Cancer Clinical Trials: A Commonsense Guide to Experimental Cancer Therapies and Clinical Trials.” (ISBN-10: 098232197X, ISBN-13: 978-0982321973; DiaMedica, New York, NY, May 22, 2012) We plan to use these findings to power a larger intervention study designed to improve enrollment in clinical trials.

Methods

Oregon Health & Science University's (OHSU) Institutional Review Board (IRB #8391) approved all study activities. The study population included patients aged 18 to 80 years with solid tumors who were seen at OHSU's hematology and oncology clinics and who had never participated in cancer clinical trials. Patients were identified by their physicians who informed them about the study, and then designated study personnel formally consented patients who agreed to take part. A randomization table was provided so study staff would know which participants would be to the Early Group and the Wait Listed Control Group (Late Group).

The study design and schedule of patient activities is included in Figure 1. Patients who consented to take part were randomized to receive the book, “Cancer Clinical Trials: A Commonsense Guide to Experimental Cancer Therapies and Clinical Trials,” in an Early Group, or a Wait-listed Control Group (Late Group). Participants in the Late Group served as controls for the Early Study Group, but all participants received the book at no cost as part of their participation in the study.

Figure 1. Pilot Test Study Design.

An oncologist with extensive clinical research experience (author TMB) and a cancer survivor who had participated in several clinical trials jointly developed the book tested in this study. It was designed to provide basic background on cancer and cancer treatment, introduce the principles that govern clinical trials, review the various types of clinical trials, highlight notable clinical trials from the past, provide practical information about the experience of participating in clinical trials, and provide basic information about cancer drugs. The contents of the book reflect the best judgment of the authors about the breadth and depth of content that cancer patients should know to make fully informed decisions about clinical trial participation and to navigate through the participation process.

Study measures included assessing demographic characteristics of participants (age, sex, marital status, educational attainment, and cancer type, date of diagnosis, and previous cancer treatment history), feasibility or the proportion of patients who read either part or all of the 160 page book, information seeking behavior, the proportion of patients who read either all or part of the book (feasibility), perceptions of the usefulness of the book (usability/acceptability), and the extent to which the book addressed gaps knowledge and attitudes about clinical trials. Study measures were administered at baseline (prior to randomization) and at Week 3 and Week 6 via either self-administration by mail or, if the subject was not able to return by mail or self-administer the survey, by study personnel over the phone (Figure 1).

Data Analysis

Analytic comparisons were made between the baseline period and the two follow-up periods. Because of the exploratory (pilot test) nature of this study, statistical power was not calculated to assess specific study outcomes. Rather, the study was designed to allow for determining effect sizes for a larger study. Thus, descriptive statistics were used to assess patient characteristics and differences in knowledge and attitudes among patients assigned to the two study groups at baseline, Week 3 and Week 6. For measures of usefulness, we collapsed Moderately to Very Useful into a single category. For only one variable, a single participant noted the chapter was ‘not at all useful’ (noted in Table legend). All other responses fell into the ‘somewhat useful’ category.

For content knowledge variables, we calculated the percent of variables scored that were answered correctly (based on 0-100%). We then used analysis of variance to compare data between the control and intervention groups at the three time periods. We used pair-wise t-tests to assess within group differences in content knowledge scores between baseline and Week 3, baseline and Week 6, and Week 3 and Week 6 for participants within study groups. For global knowledge, we calculated mean scores using a Likert scale and compared differences between participants in study groups at baseline, week 3 and week 6, using one-way analysis of variance. We used an intent-to-treat analysis approach, with the majority of statistical tests done according to study group assignment regardless of whether participants actually read the book. All tests were two-tailed, with alpha set at 0.05.

Results

Study Participants and Feasibility

Twenty patients were initially enrolled in the pilot study, and their characteristics were similar relative to study group assignment (Table 1). Mean age was in the mid to late 50's, and the majority of patients in both groups had prostate cancer. Data capture at baseline was 100% (Figure 2); at Week 3, it was 90.9% in the Early Group and 77.8% in the Late group; at Week 6, 81.8% of data was captured in the Early Group and 77.8% was captured in the Late Group. Between 60 and 85.7% of participants reported reading the entire book, according to study group assignment (Table 2). Another 11.1 to 20% read parts of the book (between 1 and 5 chapters), and 20% of participants assigned to the early group did not read any of it, while all participants in the late group read at least part of it.

Table 1. Characteristics of Study Participants.

Characteristic	Early Group (n=11)	Late Group (n=9)
Mean Age (SD)* (Range)	56.2 (15.7) (29-79)	59.9 (9.1) (50-81)
% Male	90.9%	66.7%
Marital Status
Single	27.3%	11.1%
Married/Partnered	63.6%	66.7%
Divorced	9.1%	11.1%
Widowed	0	11.1%
Education
High School Graduate	9.1%	11.1%
Some College	27.3%	22.2%
College Graduate	63.6%	44.4%
Some Graduate School	0	11.1%
Masters or Doctorate	0	11.1%
Race
White	90.9%	100%
Mixed	9.1%	0
Ethnicity Hispanic	9.1%	0
CA Diagnosis
Prostate Cancer	45.5%	55.6%
Colon Cancer	18.2%	0
Breast Cancer	0	33.3%
Testicular Cancer	9.1%	0
Pancreatic Cancer	9.1%	0
Soft Tissue Sarcoma	9.1%	0
Mucinuous Cystaadenoma	9.1%	0
Esothageal Adenocarcinoma	0	11.1%
Date of Diagnosis
Within last year	36.4%	11.1%
2-5 years ago	36.4%	44.4%
6-15 years ago	27.3%	44.4%
Types of Treatments Received
Surgery only	9.1%	22.2%
Chemotherapy only	9.1%	0
Hormonal Therapy only	9.1%	0
Surgery and Chemotherapy	27.3%	11.1%
Surgery and Hormonal Therapy	9.1%	11.1%
RT and Hormonal Therapy	9.1%	0
Surgery, Chemo, and RT	9.1%	33.3%
Surgery, RT and Hormonal Therapy	9.1%	22.2%

^*Standard Deviation

Figure 2. Data Capture* (Feasibility).

**Some survey questions were not completed, so final response rates vary according to survey question.

Table 2. Reading Behavior & Usefulness of the Clinical Trial Book at Week 3 (Early Group Only) and Week 6 (both Early and Late Groups).

Chapters Read and Usefulness	Week 3 (Early Group Only) (n=10)	Week 6 Early Group (n=9)	Week 6 Late Group (n=7)
Number of Chapters Read: Entire Book	60%	66.7%	85.7%
Read Between 1 and 5 Chapters	20%	11.1%	14.3%
Read None of it	20%	20%	0
Usefulness Among Readers
Chapter 1: What is Cancer and How is it Treated?  Moderately/Very useful	(n=8) 87.5%	(n=6)* 83.4%	(n=7) 100%
Chapter 2: What is a Clinical Trial?  Moderately/Very useful	(n=8) 100%	(n=6) 100%	(n=7) 85.7%
Chapter 3: Types of Clinical Trials: Four Phases and More  Moderately/Very useful	(n=7) 100%	(n=6) 100%	(n=7) 100%
Chapter 4: Randomized Trials and Placebos  Moderately/Very useful	(n=7) 100%	(n=6) 83.4%	(n=7) 100%
Chapter 5: Clinical Trials that have Changed Cancer Care  Moderately/Very useful	(n=7) 85.7%	(n=6) 83.4%	(n=7) 100%
Chapter 6: Is There a Clinical Trial that is Right for Me?  Moderately/Very useful	(n=6) 100%	(n=6) 66.7%	(n=7) 85.7%
Chapter 7: Who is Looking Out for You and What Can You Expect?  Moderately/Very useful	(n=6) 100%	(n=6) 83.3%	(n=7) 100%
Chapter 8: Who Pays for Clinical Trials & What Do They Cost?  Moderately/Very useful	(n=6) 66.7%	(n=6) 83.3%	(n=6) 83.3%
Chapter 9: Cancer Drugs Currently in Use and Being Tested in Clinical Trials  Moderately/Very useful	(n=6) 66.7%	(n=5) 100%	(n=6) 100%
Chapter 10: The Future of Cancer Treatment and Clinical Trials: Personalized Medicine  Moderately/Very useful	(n=6)† 83.4%	(n=5) 80%	(n=6) 100%
Mean Score for Chapter Usefulness (all chapters combined)	89.0%	86.4%	95.5%

^*1 participant did not comment on quality;

^†1 participant noted this chapter was not at all useful

Usefulness

The vast majority of participants found the book moderately or very useful (89% in the Early Group at Week 3 and 95.5% in the Late Group at Week 6) (Table 2). Only one participant found a single chapter not at all useful, which was the chapter on the Future of Cancer Treatment and Clinical Trials.

Information Seeking Behaviors

Our assessment of patients' information seeking behavior revealed that when patients seek information about their disease, between 50-67% report they depend on their physicians to tell them what they need to know, and this did not change significantly over time. Between 33.3 and 77.8% of participants reported looking online for information about their disease (data now shown). Similarly, between 44.4% and 90% of participants reported seeking information from their physicians about their treatment, while 25 to 55.5% reported looking online for treatment information (data not shown).

Knowledge

Scores on the content-specific knowledge items ranged from 82.2% correct at baseline in the Early Group to 85.9% at Week 6 (Table 3a), and ranged from 79% correct to 92.3% correct at Week 6. We found no statistical differences in content-specific knowledge scores when comparing the control (Late Group) scores to the intervention (Early Group). However, our pair-wise comparisons found that there was a statistical difference between baseline and Week 6 for content-specific knowledge scores among participants in the Late Group (Table 3b) 79% vs. 92.1% (p=0.01).

Table 3. a. Percent of Correct Content Knowledge Test Scores at Baseline, Week 3, & Week 6 According to Study Group Assignment (Scaled 0-100).

	Baseline	Week 3	Week 6
Early Group	(n=11)	(n=10)	(n=9)
Mean Knowledge Score (SD)†	82.2 (11.7)	85.3 (9.8)	85.9 (7.7)
Range	(60-93.3)	(66.7-100)	(73.3-100)
Late Group	(n=9)	(n=7)	(n=7)
Mean Knowledge Score (SD)†	79.0 (10.8)	87.5 (4.6)	92.1 (3.1)
Range	(60-93.3)	(80-93.3)	(71.4-100)
p value for Comparison of Early vs. Late Groups	0.54	0.60	0.23
†SD=Standard Deviation

Table 3b. Percent of Correct Content Knowledge Test Scores (Paired) Within Study Group
Early Group	Baseline	Week 3	Week 6	P value Baseline to Week 3	P value Baseline to Week 6	P value Week 3 to Week 6
Mean	82.2	85.3	85.9	0.07	0.07	0.80
Standard Deviation	11.7	9.8	7.7
Late Group
Mean	79.0	87.5	92.1	0.12	0.01	0.28
Standard Deviation	10.8	4.6	3.1

Scores on the global knowledge questions were significantly different between intervention and control groups at Week 3 for three variables (Table 4): ‘I understand what I need to know about agreeing to be in a cancer clinical trial’, ‘I understand what I need to know about the special tests patients need to get because they are in a cancer clinical trial’, and ‘I believe I know when a cancer clinical trial is right for me’. Overall, patients expressed more positive agreement statements for perceptions about their physician's knowledge and support for clinical trials than they did for their own global knowledge about cancer clinical trials (Table 4). By Week 6, when both study groups had the opportunity to read the book, the scores between the two groups were not significantly different.

Table 4. Changes in Global Knowledge According to Study Group Assignment.

	Baseline			Week 3			Week 6
Patient Focused Knowledge Variables	Early Group Mean (SD†)	Late Group Mean (SD)	p value	Early Group Mean (SD)	Late Group Mean (SD)	p value	Early Group Mean (SD)	Late Group Mean (SD)	p value
At this point in my care, I know everything I need to know about cancer clinical trials	2.1 (0.94)	1.7 (0.87)	0.31	3.0 (0.82)	2.3 (1.33)	0.15	3.6 (0.88)	3.7 (1.02)	0.77
I understand what I need to know about agreeing to be in a cancer clinical trial	3.5 (1.2)	3.8 (1.51)	0.57	3.8 (0.92)	2.4 (1.41)	0.02	3.8 (0.83)	3.9 (1.12)	0.72
I understand what I need to know about how cancer clinical trials are paid for	2.5 (1.04)	3.0 (1.80)	0.49	3.7 (0.82)	2.9 (1.64)	0.18	3.8 (0.97)	4.4 (0.53)	0.14
I understand what I need to know about getting a study drug versus getting another cancer treatment that is at least as good as the study drug	2.9 (1.37)	1.9 (1.17)	0.09	3.5 (0.97)	2.5 (1.31)	0.08	3.9 (0.93)	3.4 (1.27)	0.42
I understand what I need to know about the special tests patients need to get because they are in a cancer clinical trial	2.5 (1.13)	2.1 (1.36)	0.45	3.7 (0.67)	2.1 (1.36)	0.005	3.8 (0.83)	3.7 (1.38)	0.91
I believe I know when a cancer clinical trial is right for me	2.9 (0.54)	2.1 (1.54)	0.12	3.7 (0.67)	2.1 (0.99)	0.001	3.3 (0.71)	3.9 (1.46)	0.36
My doctor is very knowledgeable about cancer trials	4.1 (0.83)	4.6 (0.88)	0.24	4.3 (0.95)	4.5 (0.76)	0.63	4.1 (0.93)	4.7 (0.76)	0.19
My doctor supports my participation in cancer trials that are right for me	4.4 (0.81)	4.7 (0.71)	0.39	3.9 (0.74)	4.3 (1.2)	0.45	4.1 (0.93)	4.3 (1.11)	0.74
My doctor knows which cancer clinical trials are right for me	3.8 (0.87)	4.4 (0.88)	0.13	3.9 (0.74)	4.4 (0.92)	0.24	3.9 (0.78)	4.6 (0.53)	0.07
I know my doctor will help me decide what kinds of treatment are right for me, including cancer clinical trials	4.3 (0.79)	4.7 (0.50)	0.21	4.4 (0.84)	4.8 (0.46)	0.31	4.4 (0.73)	4.7 (0.49)	0.41

Likert Scale: 1=Strongly Disagree; 2=Moderately Disagree; 3=Neutral; 4=Moderately Agree; 5=Strongly Agree

^†SD=Standard Deviation

Discussion

This pilot study achieved its objective of assessing the feasibility and acceptability of providing a 160-page book to cancer patients to educate them about cancer clinical trials. The majority of patients who received the book read all or parts of it (>60%), and found it to be very useful (≥89%). In terms of feasibility, 75% of participants completed all measurement activities even though a few participants did not read any part of the book. This suggests a high level of commitment to taking part in educational research among cancer patients. There are many reasons why participants might not start or finish this or other similar books, though we did not specifically ask our participants about this. Such reasons could include finding time or avoidance due to anxiety related to reading about cancer treatment. This is important because the contents and delivery of patient education, especially for those with cancer, can be a black box. It is understudied, cancer patients' outcomes are often variable, and educational content often needs to be tailored to specific patient situations. This would certainly be an area for future research.

In terms of educational impact, receipt of the book resulted in significant increases in global knowledge for three variables, even in this small pilot study that included only 20 patients. Patients in the Early Group scored significantly higher on variables related to the decision to be in a clinical trial compared to those in the comparison group at Week 3. Variables related to how cancer trials are paid for and the difference between getting the study drug versus other cancer agents did not differ.

This finding leads us to conclude that the information provided in the book was especially helpful in educating patients in a way that fostered decision-making. All the participants in this study were patients with cancer who had never been in a prior clinical trial, so they had no personal experiences that would have influenced their responses. We found it interesting that that the global scores about perceptions of physician knowledge were very high at every measurement point, which was not significant across study groups. This likely indicates a halo effect or cognitive bias well known to be a beneficial effect of a physician or other health care provider, regardless of the therapy or procedure provided (19). It is also known as physician invincibility syndrome (19).It may also reflect the tertiary, research-oriented environment that this study was conducted in. These findings, even with our small sample size, support further study, especially to see how influential information contained in the book would be in decision making about participation in clinical trials.

We also assessed content-specific knowledge in our study and while scores increased between the intervention and control groups at 3 Weeks and were especially high between baseline and Week 6 for participants in the Late Group, this was not significant when Early Group scores were compared to scores in the Late Group. When we compared changes in content knowledge within study groups, we did find a significant difference between baseline and Week 3 in the Late Group. Those who received the book later in the study had a greater change in knowledge compared to those who received it earlier, a finding which may be related to the fact that this group was most likely to read the entire book. The baseline scores were slightly lower in this group compared to the Early Group scores, and their scores at 6 Weeks were higher than those at Week 3 or Week 6 for the Early Group. Reasons for completely reading the book versus parts of it is an area that should be pursued in future research, using focus group or key informant interview techniques. It may be that recentness of reading the book affected the content knowledge scores. However, the content-specific knowledge scores increased at Week 3 in the Late Group, which should not have been related to receipt of the book. We learned from our surveys that patients' information seeking behavior primarily includes physician provided information or online searches for information about both disease and treatment. It may be that the participants assigned to the Late group sought information about clinical trials on line, which affected their knowledge scores at the measurement time point. If this is the case, then including the contents of the book using an on-line format might be especially beneficial to patients considering taking part in clinical trials. Unfortunately, the number of participants in this pilot test was too small to conduct subgroup analyses. This should be the subject of further research.

Most studies on participation in cancer clinical trials examine barriers, especially for subgroup populations, like different racial groups or the elderly (1, 2, 5, 7-10, 13). Less research has been conducted on interventions to improve accrual and most of these focus on racial minorities or other underserved groups (6, 14). Clearly more research needs to be conducted in this area. Creating decision aids for cancer clinical trials will be difficult without a literature base to guide patients through evidence based on their specific type of cancer and treatment history. This study, even in its pilot form, is relatively unique in prospectively evaluating a decision aid designed to support decisions regarding clinical trial participation.

The strengths of our study include our ability to enroll and collect complete data on the majority of participants at three different times during the study. Another strength is that we determined effect sizes that we can use to power a larger study with an equally discriminating study design. Weaknesses include the small sample size, which limited our ability to do subgroup analyses that would further tease out factors that influenced our findings, such as the interaction between online information seeking behavior and changes in content knowledge scores that appear to be unrelated to reading the book that comprised our intervention. Another weakness is that we did not link assignment to the book to a specific decision about actually participating in a clinical trial. We felt it important to test the feasibility and acceptability of the book before conducting a study that would assess the direct influence of the book to decisions to take part in cancer clinical trials, which we will pursue in a larger study.

In conclusion, studies testing educational interventions that involve providing published reading material to patients with cancer are both feasible and acceptable. Providing information to patients about cancer clinical trials using a book designed for patients with cancer can influence knowledge related to the decision to participate in clinical trials.