Using Technology to Enhance
Clinical Trial Accrual
Clinical Trial Accrual
SWOG Spring Meeting
Neal J. Meropol, MD
Neal J. Meropol, MD
Chief, Division of Hematology and Oncology
University Hospitals Case Medical Center
University Hospitals Case Medical Center
Case Western Reserve University
May 2, 2014
The Problem
•
Clinical trials are the evidence base for
Clinical trials are the evidence base for
improving cancer care
•
Clinical trials represent high quality cancer
care
However,
•
Very few cancer patients take part in
clinical trials
How many patients actually take part?
y p
y
p
•
California Cancer Registry 2001‐2008
< 1%
–
< 1%
–
Al‐Refaie et al. Annals Surgery 2011
•
NCI‐Sponsored Coop Group Trials Enrollment 1996‐
•
NCI‐Sponsored Coop Group Trials Enrollment 1996‐
2002
–
1.7%
of incident cancer cases enrolled
–
Lower in racial/ethnic minorities, older patients
–
Murthy et al. JAMA 2004
•
NCI Comprehensive Cancer Centers 2012
–
14%
median
–
http://cancercenters.cancer.gov/data/sum3.html
Barriers at the Point of Care
Barriers at the Point of Care
Logistical
Patient
Physician
4Key Determinants
Key Determinants
of Accrual for Patients
•
Awareness
•
Access
•
Eligibility
•
The Decision
Influences on Clinical Trial Decision Making
Patients
Doctors
H
lth
Communities
Healthcare
Teams/Organizations
Families
Overall Goal
To optimize patient decision making
about clinical trials by improving
about clinical trials by improving
preparation for consideration of
clinical trials as a treatment option
Theoretical Model
Knowledge
Theoretical Model
Knowledge
Barriers
Attit di
l
Preparation
for Decision
M ki
Physician
Encounter
Attitudinal
Barriers
Making
Encounter
Th
The
Decision
Miller, SM; Diefenbach, M. C-SHIP: A cognitive-social health information processing
approach to cancer. In: D. Krantz & A. Baum, editors. Perspectives in behavioral medicine. Lawrence Erlbaum: New Jersey; 1998. p. 219-44.
PRE‐ACT
Survey to assess knowledge,
attitudes, and values
Automated Feedback:
Values clarification and
top barriers
Tailored Video Library
9Welcome Screen
Thank you for your interest in the PRE-ACT study. We are asking you to take part in this study because you are coming in for your first appointment with your doctor As part of this study we will because you are coming in for your first appointment with your doctor. As part of this study, we will ask you to:
• Read and approve Informed Consent documents (Informed Consent, HIPAA Authorization) • Fill out a survey (this takes about 20 minutes)
• Look at some educational materials (this takes about 15 minutes)Look at some educational materials (this takes about 15 minutes) • Fill out a second survey (this takes about 15 minutes)
Two weeks after you meet with your doctor, we will ask you to fill out one more survey. It should take about 15 minutes
about 15 minutes.
If you have any questions or concerns about this study, you can call the study team at
1-877-404-4159 or email them at [email protected]. You will see this contact information on every screen.
Sample PRE‐ACT Video Library
Preparatory Education About Clinical Trials
Mr. Doe, below is a list of video clips that you can watch to get information about clinical trials in general, and information about common misconceptions about clinical trials.
What is standard treatment?
Is there a clinical trial for everyone?
Click an item in the list to view a video. When
treatment?
Is taking part in a li i l t i l
Are there ways to deal for everyone?
you are done watching videos, click the Back
Button.
After you view the videos selected just for
clinical trial voluntary?
with transportation and financial issues?
What is How will clinical trials
videos selected just for you, you will be given the entire video library to watch at your convenience.
randomization? affect my family?
PRE ACT was compared to NCI text in a
PRE‐ACT was compared to NCI text in a
randomized multicenter clinical trial of
>1200 adult cancer patients
PRE-ACT improves:
• Knowledge
g
• Attitudes
• Preparation
More satisfaction with PRE-ACT vs. Control
12PRE ACT was compared to NCI text in a
PRE‐ACT was compared to NCI text in a
randomized multicenter clinical trial of
>1200 adult cancer patients
13 Meropol et al. ASCO 2013RESULTS
RESULTS
Demographics
Demographic
Control
PRE-ACT
Combined
Male 258 41.6% 255 41.5% 513 41.5% Female 363 58.4% 359 58.5% 722 58.5% White 544 545 1089 White 87.9% 89.1% 88.5% Non‐white 75 12.1% 67 10.9% 142 11.5% High school graduate or 148 143 291 High school graduate or less 148 23.8% 143 23.3% 291 23.6% Some college or college graduate 473 76 2% 470 76 7% 943 76 4% graduate 76.2% 76.7% 76.4% Metastatic disease 276 47.3% 254 44.8% 530 46.0% N t t ti di 307 313 621 Non‐metastatic disease 52.7% 55.2% 54.0%
Top 5 Knowledge Barriers at Baseline
Item
% Incorrect/Unsure
Most clinical trials involve a placebo (sugar Most clinical trials involve a placebo (sugar pill). 75.5% Side effects in clinical trials are usually worse 65 2% than with standard treatments. 65.2% “Standard treatments” are the best 61 9% treatments currently known for a cancer. 61.9% Informed Consent mainly protects researchers f l it 59.9% from lawsuits. 59 9% Patients in clinical trials must get their care at different places from patients getting standard 58.7% treatments.
Top 5 Attitude Barriers at Baseline
Item
% Agree
I'm afraid of the side effects I'll have on a clinical trial. 51.9% I'm worried that the treatment I'd receive on a clinical trial
I m worried that the treatment I d receive on a clinical trial
wouldn't work for me. 41.5%
I'm afraid I'll get a sugar pill (placebo) instead of real I m afraid I ll get a sugar pill (placebo) instead of real
medicine on a clinical trial. 39.3% I'm afraid that my health insurance won't pay for a clinical
I m afraid that my health insurance won t pay for a clinical
trial. 38.7%
I wouldn't ask about clinical trials unless my doctor
38 1%
brought them up first. 38.1%
Knowledge about Clinical Trials
Arm
(n)
Test
Mean
correct SD
p-value
(n)
(of 19)
p
Control
(573)
Pre
11.71
3.77
<0.0001
(573)
Post
14.28
3.78
0.0001
PRE-ACT
(505)
Pre
11.76
3.69
<0.0001
P
t
15 07
3 07
(505)
Post
15.07
3.07
Arm
Mean
SD
p-value
Arm
change
SD
p-value
Control
2.51
3.05
0.0006
PRE ACT
3 16
3 10
PRE-ACT
3.16
3.10
Attitudinal Barriers
28 items; Higher score = more agreement with barriers
Arm
T
t
Mean
SD
l
(n)
Test
(1-5)
SD
p-value
Control
Pre
2.55
0.65
<0 0001
(570)
Post
2.39
0.67
<0.0001
PRE-ACT
(504)
Pre
2.54
0.66
<0.0001
(504)
Post
2.24
0.67
A
Mean
SD
l
Arm
ea
change
SD
p-value
Control
-0.16
0.38
<0 0001
<0.0001
PRE-ACT
-0.27
0.45
Preparation for Decision Making
Higher score = greater preparedness
Higher score = greater preparedness
Arm
(n)
Test
Mean
(0-100)
SD
p-value
(n)
(0-100)
Control
(578)
Pre
72.65
15.42
<0.0001
Post
76.48
15.54
(
)
Post
76.48
15.54
PRE-ACT
(506)
Pre
72.78
15.67
<0.0001
Post
78.02
14.12
Arm
Mean
h
SD
p-value
Arm
change
SD
p value
Control
3.37
13.52
0.09
PRE ACT
4 72
12 81
PRE-ACT
4.72
12.81
PRE‐ACT Program Satisfaction
Question PRE-ACT Mean Control Mean Favors PRE-ACT l Mean (SD) Mean (SD) p-valueHow satisfied are you with the amount of
information you received? (1-5 most satisfied)
3.74 (0.77)
3.60
(0.79) 0.002 How satisfied are you with the way the
information was presented to you? (1-5 most satisfied)
3.86 (0.79)
3.65
(0.84) <0.0001 Did this program help you feel more prepared
to consider clinical trials as a way to treat your cancer? (1-5 a great deal)
3.62 (0.92)
3.43
(0.89) 0.0003 f f
Which of the following best describes your feelings about the length of this program?
(1, reasonable; 2, a little long; 3, much too long)
1.31 (0.52)
1.53
(0.63) <0.0001 Did you find this program useful for making
Did you find this program useful for making your decision about treatment for cancer? (1, yes; 2, no)
1.22 (0.41)
1.28
Conclusions
Conclusions
•
Computer‐based approaches to patient education
before the oncologist visit can improve knowledge,
attitudes, and preparation for decision making about
clinical trials Both text and tailored video were
clinical trials. Both text and tailored video were
effective.
•
PRE ACT is more effective than NCI text in improving
•
PRE‐ACT is more effective than NCI text in improving
knowledge and reducing attitudinal barriers
•
PRE‐ACT is associated with greater patient
PRE ACT is associated with greater patient
satisfaction than NCI text alone
22Barriers at the Point of Care
Barriers at the Point of Care
Logistical
Patient
Physician
23Overall Goal
To improve efficiency and accuracy of clinical
i l id
ifi
i
i
f
b
i
trial identification at point of care by automating
the extraction of clinical information and
hi
li i l i l li ibili
matching to clinical trial eligibility
Clinical Trials Identification Approaches
Clinical Trials Identification Approaches
•
CT repositories and search engines (e.g.
CT repositories and search engines (e.g.
clinicaltrials.gov)
–
Manual searchingg
–
Manual data entry
•
Patient entry database
Patient entry database
–
Critical mass of patients/studies required
•
Automated matching from electronic medical records
Automated matching from electronic medical records
Traditional Screening
Physician evaluates patient and collects data
Physician thinks about clinical trials
Access clinical trials database
Identify potential studies for disease/stage
Assess eligibility
Identify potential studies for disease/stage
Contact research staff
27Manual Screening For Clinical Trials
What is Trial Prospector?
•
An automated tool that extracts clinical and research
eligibility data, runs a matching algorithm and
displays clinical trial eligibility results at the point of
care
•
Hypothesis: Trial Prospector will
–
Be accurate
–
Save time
–
Be appealing to physicians
–
Ultimately result in improved access and accrual
to clinical trials
29Methods
•
Trial Prospector uses artificial intelligence and natural
l
i
t
t h li i l t i l li ibilit t
language processing to match clinical trial eligibility to
clinical information from multiple data sources
•
Trial Prospector is platform‐agnostic and HIPAA
•
Trial Prospector is platform‐agnostic and HIPAA
compliant, to permit future dissemination
•
Clinical criteria for matching include:
g
–
demographics, diagnosis, TNM stage, stage grouping,
metastasis, and the three most recent lab values (CBC,
renal function, liver tests, coagulation)
30Trial Prospector Components and User Access
Physician Oncology Database Secure Web‐Based Access TP User Interface Demography, Diagnosis, TNM, Stage Grouping, Metastatic Status Matching Algorithm Metastatic Status Last 3 Lab Values Eligibility C it i Lab Database Clinical Trial Database Automated Data Retrieval TP Database Values Criteria Parchman et al. ASCO Meeting Abstracts June 2013:6538 31Trial Prospector Search Screen
xxxx xxxx xxxx xxxx xxxx xxxx xxxx xxxx Parchman et al. ASCO Meeting Abstracts June 2013:6538 32Update Option xxxx xxxx p Protocol Document Eligibility g y Checklist Exclusion Report Report Parchman et al. ASCO Meeting Abstracts June 2013:6538
Use Feedback Pilot Study
GI Oncology Clinic TP Deployment 60 New Patient Visits 60 New Patient Visits
Clinical Patient E l i
Patient Specific Survey Evaluation
Summary Experience Survey Summary Experience Survey
Parchman et al. ASCO Meeting Abstracts June 2013:6538 34
User Experience
Physician Survey Results for the Trial Prospector Group Trial Prospector (60) Trial Prospector (60) Yes No Did you review a TP report for this i ? 29 (72.5%) 11 (27.5%) patient? 29 (72.5%) 11 (27.5%) To your knowledge was the information provided by TP accurate? 22 (75.9%) 7 (24.1%) Did TP save you time in identifying potential clinical trials? 16 (57.1%) 12 (42.9%) Would you recommend utilizing TP for( %) ( %) Would you recommend utilizing TP for
eligibility screening? 25 (89.3%) 3 (10.7%)
Parchman et al. ASCO Meeting Abstracts June 2013:6538 35
Summary Experience Survey Yes No Yes No Did TP allow you to spend more time with your patients? 2 (18.2%) 9 (81.8%) Did TP k i i fi d li ibl Did TP make it easier to find eligible clinical trials for your patients? 8 (72.7%) 3 (27.3%) Did TP help you communicate with 2 (18.2%) 9 (81.8%) patients about clinical trials? 2 (18.2%) 9 (81.8%) Did TP make it easier to review protocols and eligibility checklists? 9 (81.8%) 2 (18.2%) Is TP easy to use and navigate? 11 (100%) 0 Is the TP interface visibly pleasing? 10 (90.9%) 1 (9.1%) Would you recommend TP to other Physicians? 9 (81.8%) 2 (18.2%) Parchman et al. ASCO Meeting Abstracts June 2013:6538 36
Conclusions
Conclusions
•
Trial Prospector is a point‐of‐care application that is a
p
p
pp
feasible and accurate means to screen patients for
clinical trials in a busy outpatient oncology clinic
•
Physicians generally found Trial Prospector to be easy
to use and would recommend its use for clinical trial
li ibilit
i
eligibility screening
•
Program enhancements (e.g. functionality, user
interface) and further testing are underway
interface) and further testing are underway
37Study Team
Supported by NCI R01 CA127655 University Hospitals Seidman Cancer Center/Case Western – Sarah Fulton Fox Chase Cancer Center – Michael Collins (MCW) Sarah Fulton – Tyler Kinzy – Tasnuva Liu S h M i i – Brian Egleston – Linda Fleisher (CHOP) – Sharon Manne (CINJ) – Seunghee Margevicius – Neal Meropol – Dawn Miller Sharon Manne (CINJ) – Dave Poole – Suzanne Miller E i R – Mark Schluchter Karmanos Cancer Institute – Terrance Albrecht – Eric Ross – Yu‐Ning Wong Cleveland Clinic Terrance Albrecht Robert H. Lurie Comprehensive Cancer Center, NorthwesternAl Bowen Benson III
– Anne Flamm International Myeloma Foundation – Michael Katz – Al Bowen Benson, III Cancer Support Community – Joanne Buzaglo Michael Katz Fight Colorectal Cancer – Nancy Roach