Using SMART Design to Improve Symptom Management Strategies Among Cancer Patients



Status:Recruiting
Conditions:Cancer, Cancer
Therapuetic Areas:Oncology
Healthy:No
Age Range:21 - Any
Updated:9/13/2018
Start Date:May 2016
End Date:March 2020
Contact:Sarah K Brewer, MPH
Email:brewers7@msu.edu
Phone:6168024003

Use our guide to learn which trials are right for you!

In order to optimize symptom management, there must be a shift from fixed interventions "one
size fits all" to adaptive interventions tailored to individual patients. Our multistaged
interventions use the sequential multiple assignment randomized trial (SMART) design. This
design will operationalize the decision rules for switching from one intervention to another
for individual patients. Our team's work to date with breast and lung cancer patients has
shown efficacy for two home-based symptom management interventions, reflexology and
meditative practices, delivered by or with friend or family caregivers. The sustainability of
use of these two evidence based therapies and improvements in symptom outcomes will be tested
during weeks 5-8 and at the week 12 follow up of 331 patients based on power analysis,
against a control group. By sequencing two evidence based interventions, the proposed
research will impact research and practice by determining how to maximize the benefit of
symptom management during cancer treatment.

Aim 1, part a), relative effectiveness of reflexology and meditative practices. Hypothesis 1.
Patients randomized to the reflexology group will report lower severity of fatigue and lower
summed severity score from the MD Anderson symptom inventory at weeks 1-4. This hypothesis
will be tested using statistical model #1 that relates the outcome y at weeks 1-4 to the
group assignment variable (reflexology, meditative practices or control), outcome at baseline
, and other covariates (see baseline comparisons). Additional covariates will include
variables used in randomization, due to their potential impact on outcomes. If errors are
normally distributed, this model will be fit as a linear mixed effects model (LME), which
generalizes classical analysis of repeated measures. Generalized linear mixed effects (GLME)
modeling will be used with the appropriate link function and error distribution (e.g., gamma)
if the symptom severity outcome is not normally distributed and cannot be normalized using
transformations. The investigators are primarily interested in the additive effect of the
group variable, and differences in the least square (LS) means will be tested according to
the levels of variable. Aim 1, part b), characteristics of responders and non-responders.
Patients who are responders or nonresponders on fatigue will be defined as described in
measures. The characteristics of responders and their caregivers will be compared to those of
non-responders using t-tests, chi-square or Fisher's exact tests. Aim 2, Hypothesis 2.
Patients who do not respond to reflexology on fatigue during weeks 1-4 (1st intervention
stage) and have the meditative practices added during weeks 5-8 (2nd intervention stage),
will report lower severity of fatigue and improved 3 secondary outcomes: summed severity
index of other symptoms, depression and anxiety, as compared to those who are re-randomized
to continue with reflexology alone. The analytic strategy described under the analyses for
Aim 1 will be implemented for the comparison of two groups created by the second
randomization. The repeated severity measures during weeks 5-8 and week 12 will be related to
study group (reflexology alone versus reflexology and meditative practices), symptom severity
during week 4, and covariates (see preliminary analyses). The test of the significance of the
coefficient for the group variable will yield a formal test of Hypothesis 2 for the severity
of fatigue and other symptoms. PROMIS measures of depression and anxiety obtained in the week
12 interview will be analyzed using general or generalized linear models, and the test of
Hypothesis 2 for these 2 secondary outcomes will come from the significance of the
coefficient for group assignment in the second randomization. Aim 3, Hypothesis 3. Patients
who do not respond to meditative practices on fatigue during the weeks 1-4 (1st intervention
stage) and have the reflexology added during the 2nd intervention stage (weeks 5-8), will
report lower severity of fatigue and improved 3 secondary outcomes: summed severity index of
other symptoms, depression and anxiety as compared to those re-randomized to continue with
meditative practices alone. The analysis for this aim is the same as the analysis for Aim 2,
but will be performed among those who did notrespond to meditative practices during weeks
1-4. Aim 4, Hypothesis 4: Patients randomized to intervention sequences beginning with
reflexology or meditative practices will report lower severity of fatigue and improved 3
secondary outcomes: summed severity of other symptoms at weeks 1-8 and week 12, depression
and anxiety at week 12 compared to controls. The LME model described under analysis for Aim 1
will be extended to include 8 repeated measures of symptom severity (from weekly calls) and
an additional measure from week 12 interview. The test of significance of the explanatory
variable reflecting the results of the first randomization will yield a formal test of
Hypothesis 4. PROMIS depression and anxiety measures from week 12 interview will be analyzed
using generalized linear models with the following explanatory variables: group assignment at
first randomization, depression or anxiety (respectively) at baseline, and balancing
variables from randomization. Exploratory Analysis: Aim 5. To explore which dyadic
characteristics observed during weeks 1-4 are associated with optimal patient symptom
outcomes at weeks 5-8 and week 12, so as to determine additional tailoring variables for the
decision rules of selecting the first intervention stage and switching from the first
intervention stage to the second. The analyses for this aim will help build optimal
intervention sequences by determining the optimal decision rule specifying best first and
second intervention stage. This determination is not as simple as determining the best
intervention at each stage ignoring future interventions. Such simplistic approaches would
ignore longer-term effects of the intervention which was inferior at stage 1, but produced
better outcomes in a longer term if simply continued versus combined with another
intervention.

Inclusion Criteria:

- 21 year of age or older

- Solid tumor cancer diagnosis

- Able to perform basic activities of daily living (ADLs)

- Undergoing chemotherapy, hormonal therapy, or targeted therapy

- Able to speak and understand English

- Have access to a telephone

- Able to hear normal conversation

- Reporting a severity of 3 or higher on fatigue using a 0-10 standardized scale at
intake.

Exclusion Criteria:

- Diagnosis of major mental illness on the medical record (verified by the recruiter)

- Residing in a nursing home

- Bedridden

- Currently receiving reflexology or meditative practices

- Suspected or diagnosed deep vein thrombosis or painful foot neuropathy.
We found this trial at
4
sites
Tucson, Arizona 85721
(520) 621-2211
University of Arizona The University of Arizona is a premier, public research university. Established in...
?
mi
from
Tucson, AZ
Click here to add this to my saved trials
303 East Superior Street
Chicago, Illinois 60611
?
mi
from
Chicago, IL
Click here to add this to my saved trials
?
mi
from
Grand Rapids, MI
Click here to add this to my saved trials
Lansing, Michigan 48910
?
mi
from
Lansing, MI
Click here to add this to my saved trials