1. 1
Using SocialMedia (Education Blog,Twitter and You
Tube) to Increase Participation, Understanding and Use of
Diabetes Self Management Education Resources among High
Risk VA Diabetic Patients.”
Diabetes Educational Research Study
Captain James A. Lovell Federal Health Care Center,
North Chicago, IL
March 2013
Dr. Boby D. Theckedath, M.D. (P.I.) CJALFHC, Dr. Janice L.Gilden, M.D. (Co-P.I.).,
CJALFHC, Dr. Tariq Hassan, M.D. CJALFHC, , Dr. Tom Muscarello, Ph D, DePaul
University, John D. Rinkema, FISO CJALFHC, David R. Donohue, M.A., Qualitative
Technologies, Inc., Janine Stoll, R.N. CJALFHC, Rosemary Trotta, NP, JCALFHC,
NOTE:This pilot educational research project will not recommend any change or offer
advice in changing or stopping a prescribed patient’s medical therapy. In addition, we
will not be offering alternative therapies or recommend changes in MD/Clinician
prescribed regimens or treatment protocols. Our primary focus is offering and sharing
information that is educational.
Social Media and Healthcare
Technology has been changing health care for more than a century, and with each new
technological advancements—be it phone, mobile phone, e-mail, the Internet, texting,
electronic health records (EHRs), personal health records (PHRs), or social media—
(Blogs, Twitter, You Tube) there has been both celebration and apprehension.
Practitioners see the advantages of efficiency and accessibility but often feel concerned
by how each tool may overwhelm them as patients seek care or gain access to
misinformation. Privacy is also a consideration. As a result of these concerns, health
care has been among the slowest to embrace advances in communication and
information technology.
Yet technology holds the potential to improve both individual and organizational health
outcomes. As clinicians and educators, we know that increasing our patients' knowledge
about their current risk factors while facilitating collaboration with them to achieve their
health goals can improve clinical outcomes. Today patients' active use of e-mail, the
Internet, PHRs, and social media can improve access to care, enhance patient
education, facilitate screening programs, and increase adherence to treatment plans,
especially when integrated within the context of an effective provider-patient relationship.
Technology is shifting knowledge (e.g., patients can access a wealth of healthcare
2. 2
information on the Internet or seek out and use social media to share information about
care and providers, and improve their decision-making in taking control of their
healthcare outcomes.
In short, technology has many potential benefits that we, as providers, can use to
improve our communication and collaboration with patients. It can help patients learn
more about their health or medical conditions, assist in coordinating care, inform patients
about medical decisions, improve or reinforce their memory regarding instructions given
at clinic visits, increase patients' participation in care, help them learn how to cope with
disease and make health behavior changes, reduce medical errors, and, in general,
improve the quality of care patients receive. Increasing patient involvement with their
medical records may also help them take more ownership of their health care and
contribute to their health literacy.
PROJECT CHALLENGE
The Captain James A. Lovell Federal Health Care Center, North Chicago, IL in 2006 had
625 high-risk diabetes patients defined as those with an HbA1c of > 9.5% of whom 48%
either dropped out from, or did not participate in a prescribed VA diabetes self-
management education intervention program. The remaining 52% of these high-risk
patients participated by attending a one-day self-management education seminar. They
showed an overall HbA1c improvement of 1.13% in one year, and those with an HbA1c
of 9.0%, demonstrated a 3% improvement after one year. In 2010, the North Chicago VA
diabetes self management education program is representative of a 48% system wide
non-compliance or drop out rate among high risk (> 8.0% A1C) diabetes VA patients
across the United States.
We believe new communication channels that include social media ( diabetes focused
education BLOG and Twitter) are needed to engage, educate and support VA diabetes
education and treatment protocols that can successfully engage non-compliant high risk
VA diabetes patients (> 8.0% HbA1c level) to become more actively engaged and
knowledgeable about their diabetes management challenges.
(1) RESEARCH OBJECTIVES
(a) Immediate and Long-termObjectives:
The immediate objective of our this research study is to measure the effectiveness of
using social media (diabetes education BLOG and Twitter) as a new communication
channel for VA diabetic patient education to improve their HbA1c control and other
factors in a high risk non-compliant population ( >8.0% A1c). Moreover, we will
demonstrate that this goal can be achieved with modest ongoing support from VA health
professionals, and relying on an intervention group of VA diabetic peer members to
discuss their diabetic treatment challenges and their successes on an
educational/communication BLOG site. Finally, we will demonstrate the benefits of using
social media to achieve customize healthcare educational solutions that will actively
engage the VA patient in setting and achieving personalized health care plan objectives.
3. 3
In the longer term, we will work to increase the capacity of VHA to engage with VA
chronic disease patient groups for the benefit of veterans in general, including veterans
who do not receive their healthcare primarily through VHA. We will work to
institutionalize limited support from the VHA for this and similar interventions. Finally, we
will lay the groundwork to apply this educational and communication intervention to other
VA health services-provider partnerships and to VA patients with other chronic diseases.
(b) Hypotheses and Specific Aims:
D1) Diabetes persons randomized to a peer-led social media support group (SMSG) will
have a significant decrease in HbA1c when compared to persons who receive the
traditional classroom VA diabetes self management education program.
Specific Aim 1: We will determine whether individuals with high HbA1c levels (>
9.0 % A1C level) will have improved (A1c) control if they are enrolled in a peer-led
social media support group (SMSG Blog, Twitter and You Tube) that encourages
discussion about self-management strategies and their challenges in maintaining
their individual diabetes compliance level.
D2) Diabetes persons randomized to a SMSG will report greater improvements in
specific factors known to be related to HbA1c control.
Specific Aim 2: To determine the effectiveness of SMSG for improving five
specific clinical performance factors known or thought to be related to enhanced
A1c control: lipid levels, blood pressure, smoking cessation, body weight
management, and foot and eye exam status, perceived social support, involvement
in care decisions, and knowledge of personal DSME goals and A1c control.
D3) Persons in both intervention and control groups who are controlled to consensus
A1c goals will report greater involvement in care and knowledge of personal A1c
treatment, goal A1c level, and whether they are controlled.
Specific Aim 3: To examine the relationship of A1C control to self reported
involvement in care, knowledge about personal DSME treatment and A1C goal,
and awareness of A1C control.
D4) Certain aspects of the intervention will be easier to accomplish and will have greater
impact on A1C control than other aspects.
Specific Aim 4: To gather qualitative data on participant and peer-group
discussions from the social media (Educational blog, Twitter and You Tube) sites
on the impact of various aspects of the intervention. We will also collect information
about personal barriers and challenges expressed from the intervention group
conversations to train future diabetes peer group leaders. Text mining analysis
software (SPSS) will measure patient social media communications to discover
what discussion topics and issues are the most/least important; blog articles they
click on, and if they feel an opinion/or not if it is important in improving their quality
of life.
(2) BACKGROUND/CONTEXT
(a) Scientific Rationale and Theoretical Framework:
Despite consensus that effective A1c control reduces morbidity and mortality, many VA
patients in the United States continue to have suboptimal A1c control. The reason for
this poor control is not clear, but attention has focused on why providers do not respond
more vigorously to inadequate A1c control. The VHA, in particular, has adopted a
number of mechanisms aimed at optimizing provider behavior. The state of the art
4. 4
electronic health record (EHR) provides VA physicians with real time alerts when
patients present with inadequate A1c control; succinct, evidence based practice
guidelines have been widely disseminated and reinforced by senior clinicians; and
clinical managers receive feedback regarding the proportion of patients achieving A1c
goals.
Nonetheless, roughly 50% of all high risk (>8.0% A1c) VA diabetes patients in
prescribed DSME educational programs either drop out or do not attend VHA classes.
Although continued attention to providers seems warranted, a patient-focused digital
blog, Twitter and You Tube interventions may yield greater positive results. It is well
established that medication adherence is suboptimal for many patients with chronic
disease. Some studies suggest that the proportion of patients who take at least 80% of
their prescribed doses of medication is less than 75%, and this proportion decreases the
longer the patient is taking the drug. Similarly, the explosive epidemic of obesity reminds
us that changes in diet and exercise are difficult to achieve and harder to maintain.
(b) Review of Literature Regarding Testing of Related Interventions:
(1) Training patients with chronic disease self-management skills can be effective,
but the best approach is not clear.
The Chronic Disease Self-Management Program (CDSMP), developed by Lorig and
colleagues, is a community-based patient self-management education course taught by
pairs of trained leaders, at least one of whom has one or more chronic conditions. In a
randomized trial, participants in CDSMP, as compared to controls, demonstrated
improvements in several self-management skills, including exercise and communication
with physicians. Moreover, CDSMP participants also reported less fatigue and disability,
as well as hospital days. (13) This program has been adopted within VA with
encouraging early results.(14, 15). Results of other studies have been less positive,
though a mean BP change of 5 mmHg was seen by Chodosh et al in a meta-analysis for
such programs targeting diabetes and hypertension. In that meta-analysis, no one
intervention characteristic was associated with a greater effect of the intervention.(16)
Intervention characteristics examined included tailoring of the intervention to the
individual, use of a group setting, use of feedback, use of psychological services, and
delivery by the patient’s usual physician. Warsi et al had similar results in a second
meta-analysis, this time suggesting an effect on SBP that “…might be compared with
dietary sodium restriction.”(17)
(2) Group clinical interventions for patients with a shared chronic condition have
been used and studied for many years with mixed results. In a very early study,
Levine and colleagues showed that persons with poorly controlled A1c and hypertension
levels, who were randomly assigned to participate in three professionally led one hour
group session aimed at increasing understanding and self-efficacy had superior A1c
adherence and blood pressure control than controls (18-20). Other studies have had
more mixed results. In a large trial of group visits for persons with diabetes conducted in
an integrated (non-VHA) healthcare system, Wagner and colleagues found group visits
appeared to improve receipt of preventive procedures, but did not improve hemoglobin
A1c levels (21). On the other hand, chronically ill HMO members randomly assigned to
participation in monthly group meetings that involved their primary care providers
(PCP’s) had fewer hospitalizations and ER visits, improved quality of life and greater
self-efficacy than those randomized to usual care (22).
5. 5
(3) Interventions to activate patients have had mixed results.
Although observational studies suggest that patients who are more actively involved in
their care have better diabetes and blood pressure control, (23) interventions designed
to activate patients have not uniformly led to better outcomes. For example, although
Greenfield et al demonstrated that a patient activation intervention delivered in the clinic
setting could improve glycemic control, a more recent and larger study using the same
approach found no effect (24, 25). Other studies by Greenfield and colleagues have
suggested that similar interventions can improve functional status in patients with
chronic pain ( 26) or blood pressure among hypertensive persons attending a free clinic,
(27) but confirmation of these results in larger studies by other groups has been lacking.
(4) Peer-led interventions may be effective promoters of behavior change.
VA peer-led interventions have long been considered attractive options for managing
hypertension and diabetes, which is so dependent on self care behaviors. (28, 29)
However, there have been no rigorous studies of the effect of such efforts on diabetes
and blood pressure control. On the other hand, randomized trial evidence of the
effectiveness of Weight Watchers, (30) on weight loss provides powerful circumstantial
evidence that group support can affect clinical outcomes, as does the widely
acknowledged effectiveness of Alcoholics Anonymous (31).
Researchers at the Medical College of Wisconsin, Milwaukee, WI have generated
randomized trial evidence that lay opinion leaders can improve sexual risk behavior
among urban gay men by peer education and influence (32). The evidence most directly
applicable to the present proposal comes from David M. Levine, MD, ScD, a professor at
Johns Hopkins School of Medicine. In NIH funded studies in various Baltimore
communities he and colleagues have been able to demonstrate improved HTN
identification, follow up and control using community health workers (CHW) (33). A
substantial literature supports the efficacy of CHW for the control of diabetes and blood
pressure in a variety of settings (34). However, the CHW in most studies were typically
paid individuals who had received more extensive training than we propose to provide in
the present pilot study. Studies using less extensively trained, volunteer CHW have been
less rigorously evaluated. In several publications regarding a NIH funded effort involving
the use of volunteer promoters (lay health advisors) to promote A1c and cardiovascular
health in the Latino community, results were considered encouraging, but were far from
definitive (35, 36).
First, by targeting disease management skills in addition to A1c specific knowledge, we
follow Lorig’s proven example. She has argued that the mixed evidence regarding the
effectiveness of self-management reflects the emphasis on disease specific knowledge,
rather than improving self efficacy ( 37). Second, we again follow Lorig by addressing
social and emotional aspects of disease in addition to clinical issues (38). Third, we have
the advantage of the considerable resources in the Lovell Federal Healthcare Center in
North Chicago, IL where the VA health care infrastructure of regularly scheduled health
care appointments and meetings among an established group of diabetes peers, most of
whom already have the condition being addressed, as opposed to relying on fear of the
potential consequences of poor A1c control.
(5) The multi-factorial nature of our intervention is consistent with prior literature
on enhancing A1c and blood pressure control.
6. 6
A systematic review of interventions aimed at improving A1c, including blood pressure
control determined that multi-factorial interventions were most effective (39). Although
some of the categories included are directed at the physician or healthcare system; our
intervention does include both education and self monitoring, from among the categories
listed, as well as a number of other approaches. A systematic review of interventions
directed at adherence to diabetes and BP medications found that reducing the number
of pills taken was the most strongly supported intervention, but that motivational and
“complex” approaches had some evidence to support them (40). Moreover, the VA peer
group nature of our intervention is likely to be a potent motivator for improved treatment
adherence.
(c) Ongoing VHA-funded Research at Other Institutions:
VHA has funded a number of randomized trials aimed at improving A1c control by
educating patients. Kressin et al conducted a randomized trial to determine whether
adding provider-based patient education to computerized reminders had a greater effect
on A1c control than computerized reminders alone (or usual care). Although there were
interesting interactions with patient race, preliminary results indicated no clinically
significant differences among the three arms of the study. Bosworth et al have presented
results demonstrating that bimonthly nurse telephone support to patients with
inadequate A1c and BP control can control. Follow up will determine whether this
improvement persists. In another VA funded study presented in abstract form at the
2007 VA HSRD national meeting, Goldstein and colleagues have found that group
medical visits contributed to both A1c control and medication adherence. A community
delivered intervention is an important complement to these promising approaches. First,
by developing a VA patient network of socially connected individuals, the positive
benefits of VA peer education and support can be identified. Second, because this is tied
to a large healthcare system, it is more likely to be relevant to the large number of
individuals who for various reasons use multiple systems, either synchronously or
sequentially. Third, even telephone-based and group mechanisms require substantial
provider time costs that persist over time. Finally, such interventions may help to reduce
the incidents of passive chronic disease patients (29, 41).
(3) SIGNIFICANCE- Milliman Study
(a) Uncontrolled Type-2 diabetes is a Major Cause of Death and Disability:
Nationally, Improving A1c, blood pressure and cholesterol control could save billions in
healthcare dollars. Milliman, Inc., a premier global consulting and actuarial firm released
in 2010 an analysis detailing the economic burden of type 2-diabetes on major
healthcare payers and quantifying the savings that could result from better patient
management practices. This study, entitled "Improved Management Can Help Reduce
the Burden of Type- 2 Diabetes: A 20-Year Actuarial Projection," was unveiled at the
National Conference on Diabetes in Washington, D.C on April, 18th
, 2010. "The expected
growth of type-2 diabetes in America and the resulting healthcare costs are alarming,"
said Kathryn Fitch, a co-author of the study and principal and healthcare management
consultant at Milliman. "We calculated that even modest improvements in diabetes
control measures could reduce health complications, deaths and spiraling health care
costs, particularly for the elderly."
7. 7
Milliman's study estimated the impact of improving blood glucose, blood pressure and
cholesterol control in type-2 diabetes patients. According to the report, less than two-
thirds of patients meet the target ranges for any of these three measures (A1C
<7%:49%; Systolic blood pressure <130:60%; LDL <100:39%) (1). The study found that
reducing by half the number of people who are not meeting targets could, by 2031,
reduce annual costs from diabetes-related complications by nearly $200 billion, (2),
reduce diabetes-related complications by 18 percent and reduce deaths from diabetes-
related complications by 9 percent. (3)
The study noted that the diabetes epidemic will continue to expand, and improving
treatment and management practices are vital to reversing this trend. Over the next 10
years, type-2 diabetes cases outpace the growth of the U.S. population, to eventually
affect 32 million patients (14.6 percent of the population) (4). With this jump in type-2
diabetes prevalence – and with people who have the disease expected to account for 15
percent of all national healthcare expenditures by 2031, (5) – better patient management
practices are urgently needed.
The study found that, with our aging society, the elderly will be particularly affected by
the diabetes epidemic. The share of those covered by Medicare could climb by seven
percent to encompass 44 percent of all type-2 diabetes cases (6). However, if a 30
percent reduction in people not meeting targets for blood glucose, blood pressure and
cholesterol levels is achieved by 2031, Medicare could save $47 billion annually(7).
About Milliman: The firm is among the world's largest independent and actuarial
consulting firms. Founded in 1947 as Milliman & Robertson, the company currently has
52 offices in key countries worldwide. Milliman employs over 2,400 people, including
specialists ranging from clinicians to economists. The firm has consulting practices in
healthcare, employee benefits, property & casualty insurance, life insurance and
financial services. Milliman serves the full spectrum of business, financial, government,
union, education and nonprofit organizations. The study was commissioned by sanofi-
aventis, U.S. (1) Kathryn Fitch, Kosuke Iwasaki and Bruce Pyenson. "Better
Management Can Help Reduce the Burden of Type 2 Diabetes: A 20-Year Actuarial
Projection." Milliman, Inc. New York: March 2010. page 1, table 1.
(b) How our proposed research will extend knowledge and contribute to improved
quality of VA healthcare:
Many studies have found that educational interventions delivered by healthcare
professionals or highly trained CHW can improve A1c control. However, there are also a
substantial number of studies that demonstrate little or no effect of apparently similar
interventions. The VHA is the ideal setting for the present study because of the natural
connection to other community based organizations, such as local veteran groups such
as the VFW’s and American Legion posts. Moreover, the results would be of particular
interest to VHA, where provider and patient behavior has been a focus of considerable
effort, especially the importance of treating more women veterans.
Clearly, this pilot research will be of interest to other healthcare systems as they struggle
with the data supporting the importance of patient education, and the ongoing costs of
maintaining a professionally led, educational program that is typically an add-on to the
primary care infrastructure already in place. Finally, in the future community
organizations with a focus on diabetes and hypertension such as the American Diabetes
8. 8
Association could likely to be key partners and others in the dissemination of our results
from the largely white, male, participants in the proposed pilot study to other high risk
groups, such as African Americans and women. Moreover, organizations such as the
American Heart Association are likely to recognize that digital educational interventions
could be similarly effective for other chronic conditions.
(4) METHODS
(a) Overview:
Many organizations approach (SC) social computing as a list of technologies to be
deployed as needed — a blog here, a Face Book and Twitter community there — to
achieve a particular goal. But a more strategic coherent approach is to start with a target
audience and determine what kind of relationship you want to build with them, based on
what they are ready for.
Digital media and the Internet have changed the way people manage healthcare
decisions. Now more than ever, consumers (patients) are turning to the Web to research
health information, prescription and over-the-counter drugs, medical conditions, and
treatment decisions. As online support groups have created an unprecedented source
of information for patients and healthcare organizations alike - what people are finding,
learning, and talking about online are having a significant impact on their perceptions
and healthcare decisions.
In addition, there has been a lot of healthy discussion lately about supporting
professional medical boundaries with patients when using social networking (SN) and
social media (SM) technologies, such as a focus directed towards (diabetes educational)
Blog and Twitter.
Conversely, we are a huge advocate of healthcare organizations applying different social
media tools to advance professional education, collaborative communication and patient
care. When used appropriately with monitoring and moderation processes in place,
research does support the use of such technologies to raise awareness about diseases,
provide support, advance professional competencies, and improve adherence to
prescribed treatment regimes.
Two primary focus areas in this pilot research study.
#1: The Captain James A. Lovell Federal Health Care Center, North Chicago, IL will
develop a social media research focus that follows the principles of community-based
participatory research by using VA peer-led social media content targeting high risk (>
8.0% A1c level) VA diabetic patients to help them improve their A1c control. Intervention
participants at the site will have access to a research diabetes educational BLOG: A blog
(a contraction of the term "web log") is a type of website, usually maintained by an
individual with regular entries of commentary, descriptions of events, or other material
such as graphics or video. Entries are commonly displayed in reverse-chronological
order. "Blog" can also be used as a verb, meaning to maintain or add content to a blog.
In our diabetes research the content for the blog will be focused upon diabetes
9. 9
education and information that is VHA approved. Blog participants will have the ability to
post comments, share ideas and ask questions on the blog site.
Our VA diabetes blogs will also provide commentary or news on some aspect of
diabetes treatment approved by the VHA, as well as, news and information on
maintaining a healthy lifestyle. In addition, our VA diabetes research blogs will combines
text, images, and links to other blogs, Web pages, and other media related to diabetes
care and maintaining a healthy lifestyle for our intervention diabetic patient group. We
will not recommend any changes or advise in the area of changing or stopping a
prescribed therapy. In addition, we will not be offering alternative therapies or
recommend changes in MD prescribed regimens or treatment protocols.
#2: The intervention participants will have the ability to leave comments and talk about
their personal experiences and challenges in managing their diabetes, through BLOG’s
chat platforms and digital media. Twitter will provide study participants fast text
messaging to share ideas and comments. This interactive format will be an important
part of this research. We can deploy text mining software to scan the comments of study
participants to find out what topics are the most/least important to study participants.
Most blogs are primarily textual, we intent our diabetes blogs to be multi- media driven
that will include text, graphics, video, audio and podcasting on some aspect of diabetes
educational information and news approved by the VHA and other healthcare
organizations such as the American Diabetes Association.
(b) Data and Information Extraction from VA Electronic Health Records (EHR)
Data and information extraction for the intervention and control diabetes groups will be
delivered over a 12 month period. At two months, after the diabetes educational BLOG is
launched, we will do a first benchmark data extraction of intervention group patients from
VA electronic medical records to capture a measurement of current A1c level, lipid,
blood pressure, smoking cessation, body weight management, and foot and eye exam
status. We will again at months 4, 6, 8 10 and12 obtain the same data regarding the six
clinical performance measurements. If any data extraction point is missing for a patient,
we will use the most current data available.
NOTE:Participants from the intervention and control groups at the beginning of the
study will be given one Newest Vital Sign Health Literacy Test at their first available
onsite clinic appointment at the Captain James A. Lovell Federal Healthcare Center,
North Chicago, IL. In addition, participants in the study will be asked to take two onsite
Q-sort surveys; one at the beginning, the other at the end of the study during their
routine healthcare visits at the Lovell Federal Healthcare facility. See attached files
1. Newest Vital Sign Health Literacy Instrument 081211
2. Q-Sort Survey 081211
In addition, the primary goal of diabetes communication/education is to provide
knowledge and skill training, as well as to help VA patients identify barriers, facilitate
10. 10
problem-solving and develop coping skills to achieve effective self-care management
and behavior change that produce a positive health outcome for the patient/provider
team. We need some decision making data for VA management to plan new educational
and communication intervention designed to help reverse a 48% drop out rate among
high-risk diabetes patients not involved actively in their diabetes treatments. We will use
qualitative information (opinions, attitudes and feelings) extracted from the diabetes
research Blog site by text base data mining software to discover what are the most
important/least important issues VA diabetes patient are interested in and talking about.
We feel there is great richness in subjective information from the patients own words that
can be captured and blended with quantitative data to develop a strong patient profile.
Data and information extraction for the control diabetes group will be delivered over a
year, and participants will follow standard diabetes self management education
prescribed protocols; control group participants will have no access to the research
diabetes BLOG site. At two months, after the diabetes educational blog is launched we
will do a benchmark data extraction of control group patients from VA electronic medical
records to measure current A1c level, lipid, blood pressure, smoking cessation, body
weight management, and foot and eye exam status. We will again at months 4, 6, 8 10
and 12 months obtain the same follow up data regarding the six clinical performance
measurements. If any date extraction point is missing for a patient, we will use the most
current data available. We will make every effort to capture any missing patient data,
such as a A1c level from the patient before the end of the study.
(c) STUDY VARIABLES
The diabetes self-management regimen is one of the most challenging of any for chronic
illness. Patients often must perform self-monitoring of blood glucose, manage multiple
medications, visit multiple providers, maintain foot hygiene, adhere to diet and meal
plans, and engage in an exercise program. Patients also must be able to identify when
they are having problems across these functions and effectively problem-solve to divert
crises. Diabetes outcomes may be especially sensitive to problems involving literacy,
communications, understanding the importance of HbA1c control and self-management
education.
The VA Medical System has the largest and most comprehensive digital patient record
systems in the world. Regular testing of HbA1c values is now the principal way to
measure and track glycemic control in diabetic patients. Because of its importance, as a
marker of disease control, it makes sense that patient knowledge of recent and target
HbA1c values might be a useful precondition for involvement in diabetes management
and education. HbA1c variables will be extremely important to our study, in relationship
to health literacy, education and most importantly identifying socio-demographic
variables that impacts positive patient behavior.
Six clinical performance measurements, extracted from electronic medical
records
1. A1c Level (maintaining a target level) Primary
2. Lipid level management
3. Eye and Foot exam
4. Blood Pressure level Management
11. 11
5. Smoking Cessation
6. Body Weight Measurement
Additional metrics include:
Participant four digit code number (A= Intervention B= Test
Age
Race/ethnic origin
Years with diabetes
Sex
Education level
Income
Agent Orange exposure: Yes/No
Family history of diabetes? Yes () No ()
What type of diabetes do you have? Type 1 () or 2 ()
Types of diabetes medicine taken: Insulin () Insulin and oral medications () Oral
Medications only ()
(d.) Social Media Study Participant Information
Items 1 through 9 are data stored in the VA data system, which will be used for analytic
purposes during the final evaluation phase. This data should be stored and invariant. (It
is possible that items 5 and 6 could change during the course of the study.) It will not be
asked of subjects during their visits during the study. It will be accessed under Subject
ID #, assigned by VA information security officer, which will not have any personally
identifying information. Research Subject ID will be assigned (intervention) A group or
(test) B group with three numerical numbers: example A123 or B234.
Diabetes Patient Participant Profile Information
1. Age
2. Race/ethnic origin
3. Years with diabetes
4. Sex
5. Education level
6. Income
7. Family history of diabetes? Yes_____ No______ If Yes, how many
years?__________
8. Agent Orange exposure? Yes______ No________
9. What type of diabetes do you have? Type 1______ Type 2______
Items 10 through 16 will be entered on data sheet at each visit by the nurse/educator or
physician or clinician. These are pertinent clinical performance measurement findings or
evaluation at each visit.
12. 12
10. HbA1c level
11. Blood Pressure Measurement
12. Weight/BMI Measurement
13. Cholesterol level: LDL and HDL levels
14. Foot and Eye check
15. Smoking Cessation Yes_________ No________ Date__________
16. Types of diabetes medicine your taking?: Insulin:____ Insulin and oral
medications___ Oral Medications only:________
Additional Metrics include: Qualitative subject information & data
17. Patient has completed a Newest Vital Sign health literacy test (see sample)
Yes___ No___
18. Diabetes care self-efficacy assessment Yes____ No_____
19. Diabetes education training and materials covered Yes____ No_____
20. Patient correctly understood last HbA1c value
Item 17 is a data variable describing patient Health Literacy level as measured by
the Newest Vital Sign test. We will follow the standard protocol for administering
and grading the short test as described in the literature. Scores are Low Literacy,
Low Literacy Likely, and Literate. This Health Literacy level will be measured at
beginning and end of study time frame.
Health Literacy Test: Newest Vital Sign ® NVS (low literacy, low literacy possible,
literate) as assessed at study start
Items 18 through 20 are data variables which qualitatively describe the patients self-
management attitudes and attitudes toward use of technology. These attitudinal
descriptors will be assessed using Q-Sort instruments. These values will be captured at
the beginning and end of the study. Patient self-management attitudes factor group,
socio-demographic variables such as fear, stress and worry about their disease state.
18. through 20. Patient attitudes toward use of communication and social network
technology tools
13. 13
“Using Social Media (Education Blog, You Tube & Twitter) to Increase
Participation, Understanding and Use of Diabetes Self Management Education
Resources among High Risk VA Diabetic Patients.”
Capt. James A. Lovell Federal Health Care Center, North Chicago, IL 60031
To be filled out by Clinician/educator, physician or research partner at each patient visit.
Clinician or Researcher ID# _____________________
Subject ID# A or B Group ___________ Date______________
1. HbA1c __________ Date of Exam________
2. Blood Pressure :__________ Date of Exam________
3. Body Weight __________ BMI_______ Date of Exam________
4. Cholesterol Level: LDL __________ HDL__________ Date of Exam________
5. Foot and Eye Examination check: Yes_____ No_____ Date of
Exam____________
6. Smoking Cessation: Yes________ No________ Date__________
7. Types of diabetes medicine your taking?: Insulin:____ Insulin and oral
medications___ Oral Medications only:________
Qualitative Information and Data
8. Patient correctly understood last HbA1c value? Yes______ No____ Date of
Exam________
9. Patient understood biomedical assessment of their diabetes status? Yes___
No__ Date____
10. Diabetes care self-efficacy assessment? Yes______ No____ Date _________
11. Diabetes education training and treatment materials covered. Yes ___ No____
Date _____
16. 16
(f.) Q-Sort Survey
Q-Sort Survey
SAMPLE Q-STATEMENTS:(EXAMPLES)
I am afraid of my diabetes.
Diabetes education is very important to me.
I feel satisfied with my life in general.
My VA diabetes education is useful.
I am able to keep my blood sugar in control.
I find it hard to exercise, follow a diet plan and take meds.
I have trouble reading and understanding written diabetes care instructions.
I feel unhappy and depressed because of my diabetes.
My family or friends help and support me a lot in my diabetes care
I like to keep in regular touch with my VA care-givers .
I like to keep in regular touch with my family and friends.
I can’t understand the doctor or nurse spoken instructions.
I know the importance of my blood sugar level.
I feel in control of my diabetes.
I would like to use a computer or cell phone to communicate with my VA caregivers.
It is hard to use computer and email.
I would like to communicate with other VA diabetes patients for support and information.
Q-SORT PROCESS:
Data = Q-sort (A respondent constructed representation of attitudes about the subjective
topic in the context established by the researcher)
17. 17
Uses
◘ Clinical uses of individual Q-sorts as a guide for structuring follow-up interviews with
respondents
◘ Assessment of interpersonal skills
o empathy and sensitivity to needs
o development of communication and education skills
o basis of comparison of subject's self perception of interpersonal skills with the
perception of other groups of like and unlike patients
◘ Data reduction tool for collection of many Q-sorts = person-person factor analysis
◘ Generates factor space and permits inductive interpretations
◘ Useful in structural analysis of subjectivity
◘ Become the basis for data analysis of multi-respondent (extensive) R-method studies
VARIABLES
Requirements for conduct of the study require that data be:
Contextually relevant
Responses/statements in subjects’ “own words”
Uninfluenced by researchers’ own views
Unconstrained by theoretical framework
Unrestricted by constraints of multiple choices, true/false, rating scales
A. Standard statistical research measures compare by individual items in which the
variables are the individual items at question.
B. We propose to compare the subject’s attitudinal response to the research topic with
those of all other subjects. In this case, we will use the individual subject as the unit of
measure. We then use each question as contributing to the subject’s attitude. Analysis is
done using tools such as SPSS Factor Rotation and analysis.
C. The individual subject is then the independent variable.
The patients’ attitudes are analyzed by mathematical factor analysis techniques. The
output of this process is a list of clusters (or factor types) with accompanying
identification of the patients comprising membership of each group. These factor types
Condition ofInstruction
Q-SORT
EXECUTION
CONCOURSE
Q-SAMPLE
18. 18
can be viewed as dependent variables. What factor analysis does is this: it takes
thousands and potentially millions of measurements and qualitative observations and
resolves them into distinct patterns of occurrence. It makes explicit and more precise the
building of fact-linkages going on continuously in the human mind.
See attached file: Q-sort Survey Instrument
(5) DATA ANALYSIS- Descriptive Statistics will be used on analysis of measured
variable values.
A paired sample t Test will first be run. The Mann Whitney U Test will also be performed.
Results of both will be examined for significant findings for the potential differences
between control and intervention study groups. Missing data values will be eliminated
from statistical analysis. Only entries where data values are present will be used.
Statistics calculated for each variable will include:
• mean
• sample size
• standard deviation
• standard error for the mean
• histograms
Statistics calculated for each pair of variables will include:
• correlation
• average difference in means
• t-test
• standard deviation and standard error of the mean difference
• confidence interval for the mean difference
t-Test
The standard t-test compares the means for two groups of test cases. The paired
samples t-test compares the means of two variables that represent the same group at
different times. It computes the differences between values of two variables for each
case and tests whether these averages differ from 0. If the correlation is low and the
significance value is high, the independent samples t-test may be used. In a paired
samples test, means and 95% confidence intervals are examined. If the confidence
interval for the mean difference does not contain zero, this indicates that the difference is
significant. A low value for 2-tailed significance also indicates a significant difference
between the variables.
ADDITIONAL DATA ANALYSIS
Mann-Whitney Test
This test is commonly used in medical studies to test the effect of some treatment on a
group of patients. Patients are measured at the beginning of the study, treated, and then
measured again. The general scheme is as follows:
• Initial Measurement
• Intervention
• Post Measurement
Thus, each subject is a member of two groups.
19. 19
In this study, we will measure all subjects for a number of variables initially, provide
intervention (for the TEST sub-group in the form of a Social Media Use group
intervention, for the CONTROL sub-group in the form of no change in procedure), and
then measure subjects again for measures of the variables.
The Mann-Whitney U Test is the most popular of the two-independent-samples tests. It
will test that two sampled populations are equivalent in location when values are placed
in rank-ordered lists. Observations from both groups are combined and ranked.
Calculations are done on the number of times a score from each group precedes a score
from the other group. The Mann-Whitney U statistic is the smaller of these 2 numbers.
The Mann-Whitney test measures a variable value (here HbA1c level) under two groups.
We analyze the results obtained by each Group (the grouping Variable.) Group 1 is the
control. Group 0 is the intervention group. The two groups will be examined and data
ranked.
Mann Whitney U test scores on test variables are converted to ranks independent of
membership in any groups. The mean ranks of the two groups are then compared to see
if there are any significant differences. We are interested primarily in differences
between mean ranks of the study variable values between the control and the
intervention groups.
Because ranked scores are used, the distributions do not have to be of any particular
form. The descriptive statistics on these tasks indicate the numerical representation of
the shape of the distribution. Distributions are generally asymmetrical (positively
skewed.) This militates against applying an analysis that assumes a normal distribution.
Use of the Mann Whitney test thus alleviates problems that would occur if we relied on
statistical tests that assume a normal distribution.
We will present both Mann Whitney U Test results indicating ranking relationships on
test variables, as well as raw variable values and distributions. Without the latter, a
reader would have little idea of what the raw results were like. We will present tables that
list rank data for each of the variables found to have a statistically significant test result.
The mean rank for each lists the average of the ranks for each group. Similar values
would indicate similarity in Group location. Small significance values (< .05) indicate that
the two Groups have different locations when placed in rank-ordered lists. This would
indicate a significant difference in outcome based on the difference in intervention used.
Q-sort Factor Rotation Using SPSS Software
The instrumental basis of Q-methodology is the Q-sort technique, which conventionally
involves the rank-ordering of a set of statements from agree to disagree. Usually the
statements are taken from interviews or focus groups, hence are grounded in concrete
existence; for purposes of convenience, however, the Q-sample in this study will consist
of statements taken from our previous VA pilot study and from Larson's (1984) CARE-Q
set.
This study will also use a concourse of patient statements to identify patient attitudes
toward health literacy and patient education. They will be summarized and the resulting
set of representative statements will be given to study participants for rank ordering
according to agreement with each. This will give us attitudinal profiles of each patient.
Thus, patient attitudinal status is a variable to be studied. This will be done by analyzing
the overall similarity in response patterns of each patient to all other patients. By using
20. 20
SPSS statistical analysis software, we will be able to identify the factor groups that exist
in the patient population with respect to the attitudinal measures.
Factor analysis is a method of data reduction. It does this by seeking underlying
unobservable (latent) variables that are reflected in the observed variables (manifest
variables). There are many different methods that can be used to conduct a factor
analysis (such as principal axis factor, maximum likelihood, generalized least squares,
unweighted least squares), There are also many different types of rotations that can be
done after the initial extraction of factors, including orthogonal rotations, such as varimax
and equimax, which impose the restriction that the factors cannot be correlated, and
oblique rotations, such as promax, which allow the factors to be correlated with one
another. Analysts must determine the number of factors that they want to extract. Given
the number of factor analytic techniques and options, it is not surprising that different
analysts could reach very different results analyzing the same data set. However, all
analysts are looking for simple structure. Simple structure is a pattern of results such
that each variable loads highly onto one and only one factor.
The study sample size will be sufficient to produce a stable outcome. The resulting
factors will be identified by those individual statements that load most heavily on them.
DIABETES SOCIAL MEDIAPROJECT BLOG CONTENT
Weblog Characteristics
Contemporary weblogs illustrate a number of characteristics: They are expected to be
visible to the Web at large; they comprise frequently updated commentary, articles or
dialogue as a dated chronology, most recent first; they have some form of archiving
feature, usually accessible through an onscreen calendar; are increasingly open to
commentary from visitors (creating a ‘threaded’ dialogue); and are usually free to create
and update. Most blogs are visible on the Web.
Further, a dedicated blogging environment will be based upon a content management
system (CMS) – a simple interface through which content can be updated – be
configurable through use of a programmatic interface permitting layout and look-and-feel
adjustment, have an option to accept visitor feedback commentary, and offer the ability
to include images and XML feeds (a method of viewing content indirectly).
An increasing number of blogs now incorporate syndication features such as RSS feeds.
Typically, these permit remote access of blogs by feed aggregators, programs which
monitor hundreds or thousands of blogs for updates, informing the user of fresh or edited
content. Such a simple yet powerful alert flagging system helps bloggers keep in touch
with their favorite blogs.
The Lovell Federal Health Center diabetes research social media blog content will not
feature content information that proposes a specific patient medical treatment regime or
specific patient medical course of action. Our blog’s content will contain information that
is general and educational in nature, and touch on educational content in six key areas
that includes the following:
1. HbA1c Level (maintaining a target level)
2. Lipid management
3. Eye and foot exam
4. Blood pressure management
21. 21
5. Smoking Cessation
6. Body weight measurement; importance of exercise and following a healthy diet
The blog content posted to the research site will be static in nature, in that the reader
cannot change the content. Then, there is the simple issue of the actual written content
on the blog. Nothing will be quite as important for achieving success as this. Not only
are we going to want to post new content to the blog on a regular basis, but we will
insure the content is top quality: well-written, truthful, and relevant to our blog’s topic,
such as the importance of maintaining a good HbA1c level, blood pressure monitoring,
following a healthy diet and regular exercise.
Updating the BLOG content: The blog content will be updated at least weekly
according to the directives of the PI and Co-PI. We are not planning to notify the IRB on
planned copy changes to Blog content, because we are not developing Blog content with
specific medical therapies or recommend changes in physician prescribed treatment
regimens. Our blog content will be educational in nature and if a diabetes patient has a
concern about their specific health issue we ask them to contact their medical provider
directly.
Training planned for the social media group. There will be ongoing training by the
use of conference calls. The training should include instructions on not using PHI on the
blog, the necessity for thoughtful comments (realizing the potential implications for
someone’s health regarding postings), and the necessity to check with their MD prior to
changing or stopping a prescribed therapy. No postings will be allowed that provide
links to alternative therapies or recommend changes in MD/Clinician prescribed
regimens.
We will not be using ANY (PHI) personal health information on the Blog site and each
research volunteer participant will be given a research study ID number by John
Rinkema (FISO) before having direct access to the blog site in order to have the ability
to leave personal comments on the site. There will be no postings that advocate any
diabetes therapies or specific medical prescribed regimens. We will not be giving
medical advice OR treatment instructions on this blog site.
Member of the study team will monitor the Blog postings daily. John Rinkema is the
Information Security consultant to this project and he will review all Blog contents prior to
information being posted and will monitor the Blog and its responses to ensure Privacy
and Security is maintained.
Study team members will quickly intervene or remove misleading information from the
blog. The study team is committed to ensuring optimal privacy and security of each Blog
recipients. If a study participant inadvertently places information in response to the blog
(even after being warned not to do so) the study team will remove this information as
soon as the team becomes aware of this information.
David R. Donohue will serve as the blog editor, where he will be in charge of researching
and writing content for the blog site. All subject and blog content will be reviewed by Dr.
Boby G. Theckedath, M.D. and the research study clinical team members, Janine Stoll,
R.N. and Rosemary Trotta, N.P and John Rinkema, ISO. Study research team member
will have final review of the copy content before being posted on the Blog site. The blog
22. 22
site will be monitored daily and any team research member can remove any copy from
the blog site if any adverse issues are raised on the particular blog’s content.
Adverse Event Monitoring:
Dr. Boby G. Theckedath, M.D. and Dr. Janice L. Gilden, M.D. and other research study
clinical team members including, Janine Stoll, RN and Rosemary Trotta, NP, Tom
Muscarello and David R Donohue will all monitor the BLOG for potential adverse
information.
In addition, Part III application notes SAEs will be reported within 24-hours to the IRB.
The study team will daily monitor the blog to meet this reporting requirement. Study team
members can intervene or take down any information that may be potentially harmful to
blog participants. John Rinkema has oversight of Privacy and Security issues, all study
team members can intervene and remove potentially harmful information if they deem it
necessary.
(5) SPSS Data Analysis/Text Mining the BLOG Site Content:
Use text analysis software such as, IBM’s SPSS to measure patient social media
communications to discover what discussion topics and issues are the most to least
important to the VA intervention diabetic research study patients. Valuable insight can be
extracted from what the intervention patients are saying in their comments and blog e-
mails that can be used to build cluster factor groups and identify key areas important to
the intervention patients.
Text mining, sometimes alternately referred to as text data mining, roughly equivalent to
text analytics, refers to the process of deriving high-quality information from text. High-
quality information is typically derived through the divining of patterns and trends through
means such as statistical pattern learning. Text mining usually involves the process of
structuring the input text (usually parsing, along with the addition of some derived
linguistic features and the removal of others, and subsequent insertion into a database),
deriving patterns within the structured data, and finally evaluation and interpretation of
the output. 'High quality' in text mining usually refers to some combination of relevance,
novelty, and interestingness.
Typical text mining tasks include text categorization, text clustering, concept/entity
extraction, production of granular taxonomies, sentiment analysis, document
summarization, and entity relation modeling (i.e., learning relations between named
entities).
(a) VARIABLES
Requirements for conduct of the study require that data be:
Contextually relevant
Responses/statements in subjects’ “own words”
Uninfluenced by researchers’ own views
Unconstrained by theoretical framework
Unrestricted by constraints of multiple choices, true/false, rating scales
23. 23
A. Standard statistical research measures compare by individual items in which the
variables are the individual items at question.
B. We propose to compare the subject’s attitudinal response to the research topic with
those of all other subjects in both the intervention and control groups. In this case, we
will use the individual subject as the unit of measure. We then use each question as
contributing to the subject’s attitude.
C. The individual subject is then the independent variable.
The patients’ attitudes are then analyzed by mathematical factor analysis techniques.
The output of this process is a list of clusters (or factor types) with accompanying
identification of the patients comprising membership of each group. These factor types
can be viewed as dependent variables. What factor analysis does is this: it takes
thousands and potentially millions of measurements and qualitative observations and
resolves them into distinct patterns of occurrence. It makes explicit and more precise the
building of fact-linkages going on continuously in the human mind.
(b) Pilot Research Study Size:
The Lovell Federal Healthcare Center diabetes social media study will use a total of 100
participants; 50 each for the intervention group and control group. Intervention
participants will have access to the blog educational materials and information posted on
the diabetes research blog site at North Chicago. Control group participants will not have
access to the research diabetes blog site, but will follow the standard diabetes self
management education VA protocols, now approved by the VHA.
This project will use probability and stratified sampling modeling, because it allows a
calculation of the sampling error and controls for the following factors.
1. A desire to minimize variance and sampling errors and to increase precision
2. A desire to estimate the parameters of each stratum and have a readable
sample size for each.
3. A desire to keep the sample element selection process simple.
Study subject frame will be identified and selected through electronic VHA medical
records (with an HbA1c level > 9.0%) in either 2010 and/or 2011.
♦ Be at least 18-years old
♦ Have a prescription for a glucose control medication or supplies, or one hospitalization,
or two outpatient visits with a diabetes related ICD-9 code
♦ Has seen their primary care provider (PCP) in the prior 12 months
♦ Scheduled to see the same PCP, in the next 2 to 12 months.
♦ Have digital access to the WEB or able to access the blog site from a digital portal
(public library) or a research kiosk at the North Chicago VA Medical Center.
Subjects will be contacted by in-person clinic invitation, e-mail or USPS letter to invite
them to participate in the study. The pilot research study will receive (IRB) institutional
review board approval, and written informed consent obtained from all participants.
24. 24
Cognitively impaired VA patients will not be asked to participate in this study; patient’s
who cannot read sample questions, due to poor eye sight will be included, provided
subjects can read larger type instructions.
Captain James A. Lovell Federal Healthcare Center Social Media Recruitment Plan
The recruitment letter clearly states that the study is completely voluntary, that there is
no penalty if the veteran decides not to participate and that refusing to participate will not
affect the health care provided.
Subjects will be recruited in one of three ways and invited to participate in the study
1. Diabetes Home Management Class – this class is held twice a month and open
to all Veterans. Veterans are referred to class by their Primary Care provider, but
also from specialty clinics, nutritionist and self referrals. The recruitment letter
will be read at the end of the class and a copy placed in their educational
materials folder.
2. Endocrine Clinic – Veterans with diabetes seen in the endocrine clinic will be
handed the letter by the intake nurse, not by the patient's provider to decrease
any possibility of coercion or duress.
3. Lab Reports –Hemoglobin HbA1c lab reports will be run monthly. Those
patients with HbA1c > 9% will be send the recruitment letter by USPS.
Veterans will only receive the recruitment letter once to decrease any possibility of
coercion. A list of patients sent or given the recruitment letter will be kept to ensure no
duplications. This list will be kept in a secure location under password protection. Once
the recruitment process is complete the list will be destroyed.
Rosemary Trotta, NP Janine Stoll, RN, BSN
Nurse Practitioner/Endocrine Certified Diabetes Educator
Captain James A Lovell Captain James A Lovell
Federal Health Care Center Federal Health Care Center
(224) 610-7012 (224) 610-7007
Project Participant Recruitment Letter: See VA Social Media VA Patient Recruitment
Letter 081211
(6) DATA, SAFETY AND INFORMATION SECURITY MONITORING PLAN
PROTECTION OF HUMAN SUBJECTS
This VA social media research study will follow all guidelines under the: The Federal
Policy for the Protection of Human Subjects, 38 CFR Part 16, and all guidelines outlined
by the Department of Veterans Affairs for the protection of human subjects, the Privacy
Act, and the HIPAA Privacy Rule, HITECH, and the ARRA.
25. 25
This study is not a clinical research trial. There are no physical procedures, lab tests,
drug or medical interventions done as an express part of this study. Such interventions
will be done as a part of the patients’ regular care and the results of those interventions
may be used as data variables for analysis.
The interventions specifically proposed in this study will be educational in nature and
would not be expected to have an adverse impact on physical health or social or
emotional functioning. Since there are few if any adverse effects expected related to
participation; the potential benefits of study participation on overall health and well being
can reasonably be expected to be greater than the risks.
Educational interventions will use materials that already exist in the VA to ensure that
the information is accurate and in line with accepted disease management strategies.
Participants in the study will be receiving their normal care through the diabetes clinic
and will be receiving enhanced educational and case management intervention through
the project. They will be consented using a consent form and process approved and
monitored by the Hines/Captain James Lovell Federal Health Care, Institutional Review
Board and the Hines/CJLFHC Research and Development Committee.
DATA SECURITY PROCEDURE
All health and demographic data obtained and used in this study will be de-identified in
accordance with VHA Handbook 1605.1, May 17th
, 2006, Appendix B. This data will,
then, not be considered sensitive information. It will not contain any information that will
identify the patient and thus will afford no reasonable basis to believe that the
information can be used to identify an individual.
Subject numbers will be assigned as identifiers to the physical files, with a Master Index
List also maintained for linking subject name and subject number. This list will also be
maintained in the investigator's locked filing cabinet. It will be available only to the PI,
Co-PI and the study Data Security Officer.
The research study will generate
non-identifiable, qualitative, interview data;
certain non-identifiable, clinical measures(HbA1c, blood sugar levels, weight,
height, BP); and
non-identifiable, qualitative or questionnaire data on health status.
Ongoing and summary analysis will involve the use of the data elements listed above
with certain demographic or clinical measures or elements for each patient. These will
be obtained from the VA patient clinical database using the Master Index List to obtain
any de-identified data needed to analyze research data for trends.
There will be no verbatim written transcripts of any focus groups, meetings, or patient
intervention sessions.
There will be no video or audio recording.
No research data will be stored off-site.
Analysis of data will be done by accessing the study data base.
VA Human Research Protection Procedure
The PI or Co-PI must submit Appendix C (Data Security Checklist for Principal
Investigators) and Appendix D (Principal Investigator's Certification: Storage & Security
26. 26
of VA Research Information) at the time of any new protocol submission to the IRB, and
annually, by April 15th, for each active research protocol.
The VA Central Office (VACO) has established rules for protection of data used in and
derived from research projects. These rules apply to already existing data (retrieved for
the purposes of the research), data created through the research, data repositories, and
other uses of and transfer of VA research data. Among the requirements from VACO is a
requirement for investigators to complete a “Data Security Checklist for Principal
Investigators” and “Principal Investigator’s Certification: Storage & Security of VA
Research Information” at the time of protocol submission and annually thereafter.
In compliance with VA mandates, CJLFHC has named an Information Security Officer
(ISO) who (among other Medical Center duties) sits on the IRB Committee for the
purpose of reviewing research proposals to ensure that information security
requirements have been satisfied. The ISO reviews each clinical research proposal (any
non-exempt research proposal that involves human subjects) to determine what types of
data will are involved in the research project, where and how the data will be stored or
transferred, the risks of security breaches, and other required elements.
Policies:
All human research protocols must receive ISO approval in order to be approved
by the IRB & R&D Committees.
The IRB Committee seeks the ISO’s guidance on information technology (IT)
matters.
The Research Service Office of Research Compliance (ORC) has the authority to
evaluate investigators’ compliance with their stated IT security plans.
The VHA requires an assessment of the mechanisms in place to ensure (1) security of
data and all files, (2) confidentiality of data derived from research subjects, (3) release of
data in accordance with VHA regulations and policies, and (4) control of data so that
reuse of the data is within an approved research protocol and in compliance with VHA
procedures. (VHA Handbook 1200.1, “R&D Committee Handbook”, June 16th
2009) To
that end, the PI must submit a Data Security Plan to assist in creating a data security
system that protects the privacy of subjects, is VHA and HIPAA compliant and minimizes
the risk of data breaches. This plan needs to be approved by the R&D Committee in
order for the study to be approved. The form is used to establish the use of any
identifiable patient information on study subjects. The PI on this study will be able to
certify on page 1 of the document that:
All VA sensitive research information is used and stored within the VA
Data will be identified only by subject number at the time of collection and
during all phases of data analysis
VA Research Protection Review Officers
Privacy Officer (PO)
Reviews the documentation of each human research project to ensure that
information security requirements have been satisfied and documented.
Is a non-voting member of the IRB committee).
Principal Investigator
Completes the “Data Security Checklist for Principal Investigators” and “Principal
Investigator’s Certification: Storage & Security of VA Research Information” and
submits them with the protocol submission to the RDC.
27. 27
.
Research Compliance Officer (RCO)
Conducts routine or for-cause audits of investigator practices regarding the use,
disclosure, storage, transfer and destruction of research-related PHI and III.
North Chicago VA also has a research sub-committee that conducts random
audits of consents and protocol compliance.
Transmission of Data
If data is sent to a sponsor and the data was collected under an informed consent and
HIPAA authorization permission must be obtained, the ACOS/R&D, the ISO and the
Privacy Officer. Local policy dictates that transfer or storage of all research data off-site
is subject to approval. The project will be reviewed to ensure that the protocol, consent,
and the agreement with the sponsor adequately address data security issues.
Furthermore, if the data is sent to a sponsor it must be transmitted in compliance with all
VA requirements. While the data and all copies are in VA possession, all VA
requirements must be meet.
Data Storage
Data will be entered into a computer database within one week of collection. This will
include demographic data, clinical encounter information, and Entered data will be
compiled in a single, integrated database using SQL. We will currently use data base
software on high-end Microsoft based computers for our data analyses. IBM/SPSS
software will be used for statistical analysis of data.
The database will be maintained on a single computer that is not linked to public access
servers, is maintained by the VA principal investigators, and is stored in a locked office
in a secure VA building. The database will be password protected. All patient data
records will have identifying information removed. Each patient will have a study ID that
is maintained by the PI and Co PI. Files are built using self-documenting systems, so
that codebooks are readily available. Analytical output files for key analyses are
documented, printed, and archived for audit, as necessary. Computer files will be
backed up after each use. Once the study is closed data is removed to a secure area at
CJLFHC and destroyed at a pre-determined date as specified in the protocol. (usually
retained for five years)
Digital Data
No data will be stored offsite. The use of any personal storage devices at the VA is
strictly prohibited.
Laptop computers and other computer equipment, operating systems, and software will
be purchased, setup and maintained by VA personnel, following standard VA procedure.
Appropriate VA security measures, firewalls and encryption will be used. No laptop or
other computer equipment will leave the VA premises. This is not allowed. In the event
that such removal is required, there are specific forms that must be completed in order to
take equipment off-site. If a researcher would like to take Government Furnished
Equipment (i.e., VA or VAPHS Research Foundation purchased) off-site, he will need a
Check-Out Sheet from the Research Office.
VA Patient Database Location and Project Information Storage
The Patient Database (identifiable) will reside in the ISO Special Projects
28. 28
Folder which is open only to (John Rinkema) and is located on the T-Drive/World
Directory/ISO Special Projects Folder
The De-identified Database will be located in the Research Study Folder by
Project Number. This is located on the T Drive/Research/Research Projects/(Study
Number).
The Research Blog will also be in this site in a separate folder. This site T
Drive/Research/Research Projects/(Study Number) will be accessible to the
Investigative Team and Research Management only.
We will provide accounts to the research team to access the T Drive/Research/Research
Projects/(Study Number) site. The Blog may be posted directly to the internet from this
site.
John D. Rinkema
Information Security Department Head
(005R5/ISO)
Captain James A. Lovell Federal Healthcare Center
3001 Green Bay Road
North Chicago, IL 60064
Office: 224.610.3805
FAX: 224.610.3822
Cell: 847.833.6161
john.rinkema@va.gov
information.securityofficer@va.gov
Paper/Forms
Paper data will be maintained in a locked filing cabinet in the principal investigator's or
CO-PI office. No copies of the original notes will be made. These will include the
following de-identified materials:
Master Index List as described above;
Consent and all other participation forms and documents;
Handwritten notes on focus group meetings and patient educational meetings;
Researcher summaries of focus group meetings and patient educational
meetings;
Clinicians summaries of patient clinical progress
There is an underlying risk of breach of confidentiality whenever medical records are
accessed or used in the course of medical treatment and research. To minimize the
risk of breaches of confidentiality all study personnel will complete VA privacy training,
information (cyber) security training, as well as yearly VA research education. No
identifying data will be removed from the VA computer system or the Lovell Federal
Healthcare Center campus.
Subjects will be selected randomly from the Lovell Federal Healthcare Center patient
population using only the medical selection criteria that are relevant to the study. This
procedure will maximize the likelihood of equitable subject selection.
Published Data
29. 29
Published accounts of the results of the study will protect the anonymity of the
participants by reporting only aggregate data.
POTENTIAL RISKS
The interventions in the study are educational in nature and use accepted disease
management strategies and are not expected to have physical, social, economic or
emotional impact on the subject. The risk of loss of confidentiality through participation in
the study is not significantly greater than the risk that occurs in the course of normal
medical treatment where records are accessed and reviewed in the process of providing
care.
Informed consent will be sought from each prospective subject or the subject's legally
authorized representative, in accordance with, and to the extent required by the VA.
Informed consent will be appropriately documented, in accordance with, and to the
extent required. When appropriate, the research plan makes adequate provision for
monitoring the data collected to ensure the safety of subjects.
Safety Monitoring Plan
All members of the research staff who interact with subjects will be trained to recognize
and report events that might be categorized as adverse. Procedures for reporting
suspected adverse events will be standardized and will be provided in writing to all
personnel. Specifically, adverse events will be reported within 72-hours to the on-site
medical director, the project principal investigator, and the chair of the university’s
Human Subjects Protection Committee. Reportable events will include (but will not be
limited to): (a) problems regarding the consent process, and (b) distress during the data
collection procedures or distress caused by notifying a patient about the diagnosis. The
chair of the Human Subjects Protection Committee and the principal investigator will
determine whether the event is study-related, and the principal investigator will forward
reports of study-related adverse events to the sponsor.
APPENDICES
Hines/CJLFHC Data Security forms C &D (samples; actual forms are completed at the
time of IRB submission. For this study all research data is used, stored and remains in
the VA so the review has fewer data issues that need to be addressed on these forms)
STUDY PERSONNELL:
The Part III application notes IRB training is not required for John Rinkema. The
Diabetes Patient Participant Profile Information states the subject ID is assigned by the
security officer. This security officer is then considered engaged in research and human
subjects training requirements must be met. All study members will meet the training
requirement.
Dr. Tom Muscarello, Ph. d will design and manage Q sort data collection and statistical
data analysis procedures. This will include the selection of any computer software
necessary. He will, under the guidance of the PI and Co-PI, ensuring that all data
security and privacy provisions are implemented under Federal and VA standards. He
will work with the PI, Co-PI, clinicians, and the project coordinator on data analytic and
30. 30
IT related aspects of the development of research plans, study design, and data
collection methodology, including the design of Case Report Forms.
Dr. Muscarello will also participate in designing and building of test instruments, report
writing and editing, dissemination and presentation of findings at conferences. This will
involve part time work throughout the duration of the project. All work will be done offsite
at DePaul University. Any DePaul local IRB approvals will be obtained. No extra
equipment will be needed.
David R Donohue, M.A. will research and write all Blog content for the project and
review the content with the PI, Co-PI and FISO consultant and other research partners
before posting online. He will coordinate all aspects of social media resources with the
PI, Co-PI and FISO to optimize the quality of educational and information content. He
will be under the guidance of the PI and Co-PI, to ensure that all data security and
privacy provisions are implemented and followed under current Federal and VA
standards. In addition, he will work with the P.I., Co-PI, FISO, clinicians, and other
research partners to insure that project tasks and timelines are met to insure that all
project goals are accomplished on schedule and collaborative partnerships are
maintained. He will assist in report writing, editing, dissemination and presentations of
findings at conference and other professional events.
Study personnel contact information:
Social Media Research Partner Contact Information
Dr. Tariq Hassan, M.D. (Senior Advisor) (224) 610-3701, tariq.hassan@VA.gov
Dr. Tom Muscarello, Ph.d(DePaul University) (312) 375-5573,
TMuscare@cdm.depaul.edu
Dr. Boby G. Theckedath, M.D. (224) 610-7013, Boby.Theckedath@va.gov
Dr. Janice L Gilden, M.D. Janice.Gilden@VA.gov
Janine Stoll, R.N. (224) 610-7007, Janine.Stoll@va.gov
Rosemary Trotta, NP (224) 610-7012, rosemary.trotta@va.gov
David R Donohue, M.A. Mobile (847) 651-3891, ddonohue9@gmail.com
John Rinkema, MS (224) 610-3805, john.rinkema@va.gov
REFERENCES AND FOOTNOTES:
13. Lorig KR, Sobel DS, Stewart AL, et al. Evidence suggesting that a chronic disease self-
management program can improve health status while reducing hospitalization: a
randomized trial. Med Care. 1999; 37:5-14.
14. Linnell K. Chronic disease self-management: one successful program. Nurs Econ.
2005;23:189-191.
15. Nodhturft V, Schneider JM, Hebert P, et al. Chronic disease self-management: improving
health outcomes. Nurs Clin North Am. 2000;35:507-518.
16. Chodosh J, Morton SC, Mojica W, et al. Meta-analysis: chronic disease self-management
programs for older adults. Ann Intern Med. 2005;143:427-438.
17. Warsi A, Wang PS, LaValley MP, Avorn J, Solomon DH. Self-management education
programs in chronic disease: a systematic review and methodological critique of the
literature. Arch Intern Med. 2004;164:1641-1649.
31. 31
18. Levine DM, Green LW, Deeds SG, et al. Health education for hypertensive patients.
JAMA. 1979;241:1700-1703.
19. Morisky DE, Bowler MH, Finlay JS. An educational and behavioral approach toward
increasing patient activation in diabetes and hypertension management. J Community
Health. 1982;7:171-182.
20. Morisky DE, Levine DM, Green LW, et al. Five-year blood pressure control and mortality
following health education for hypertensive patients. Am J Public Health. 1983;73:153-
162.
21. Wagner EH, Grothaus LC, Sandhu N, et al. Chronic care clinics for diabetes in primary
care: a system-wide randomized trial. Diabetes Care. 2001;24:695-700.
22. Scott JC, Conner DA, Venohr I, et al. Effectiveness of a group outpatient visit model for
chronically ill older health maintenance organization members: a 2-year randomized trial
of the cooperative health care clinic. J Am Geriatr Soc. 2004;52:1463-1470.
23. Schulman BA. Active patient orientation and outcomes in hypertensive treatment:
application of a socio-organizational perspective. Med Care. 1979;17:267-280.
24. Greenfield S, Kaplan SH, Ware JE, Jr., Yano EM, Frank HJ. Patients' participation in
medical care: effects on blood sugar control and quality of life in diabetes. J Gen Intern
Med. 1988;3:448-457.
25. Williams GC, McGregor H, Zeldman A, et al. Promoting glycemic control through
diabetes self-management: evaluating a patient activation intervention. Patient Educ
Couns. 2005;56:28-34.
26. Greenfield S, Kaplan S, Ware Jr JE. Expanding patient involvement in care: effects on
patient outcomes. Ann Intern Med. 1985;102:520-528.
27. Kaplan SH, Greenfield S, Ware JE, Jr. Assessing the effects of physician-patient
interactions on the outcomes of chronic disease. Med Care. 1989;27:S110-S127.
28. Kulcar Z. Self-help, mutual aid and chronic patients' clubs in Croatia, Yugoslavia:
discussion paper. J R Soc Med. 1991;84:288-291.
29. Nessman DG, Carnahan JE, Nugent CA. Increasing compliance. Patient-operated
hypertension groups. Arch Intern Med. 1980;140:1427-1430.
30. Heshka S, Anderson JW, Atkinson RL, et al. Weight loss with self-help compared with a
structured commercial program: a randomized trial. JAMA. 2003;289:1792-1798.
31. Krystal JH, Cramer JA, Krol WF, Kirk GF, Rosenheck RA. Naltrexone in the treatment of
alcohol dependence. N Engl J Med. 2001;345:1734-1739.
32. Kelly JA, Murphy DA, Sikkema KJ, et al. Randomised, controlled, community-level HIV-
prevention intervention for sexual-risk behaviour among homosexual men in US cities.
Community HIV Prevention Research Collaborative. Lancet. 1997;350:1500-1505.
33. Levine DM, Bone LR, Hill MN, et al. The effectiveness of a community/academic health
center partnership in decreasing the level of blood pressure in an urban African-American
population. Ethn Dis. 2003;13:354-361.
34. Brownstein JN, Bone LR, Dennison CR, et al. Community health workers as
interventionists in the prevention and control of heart disease and stroke. Am J Prev Med.
2005.
35. Balcazar H, Alvarado M, Hollen ML, Gonzalez-Cruz Y, Pedregon V. Evaluation of Salud
Para Su Corazon (Health for your Heart) -- National Council of La Raza Promotora
Outreach Program. Preventing Chronic Disease. 2005;2:A09.
36. Kim S, Koniak-Griffin D, Flaskerud JH, Guarnero PA. The impact of lay health advisors
on cardiovascular health promotion: using a community-based participatory approach. J
Cardiovasc Nurs. 2004;19:192-199.
37. Lorig KR, Holman H. Self-management education: history, definition, outcomes, and
mechanisms. Ann Behav Med. 2003;26:1-7.
38. Bodenheimer T, Lorig K, Holman H, Grumbach K. Patient self-management of chronic
disease in primary care. JAMA. 2002;288:2469-2475.
39. Fahey T, Schroeder K, Ebrahim S. Interventions used to improve control of blood
pressure in patients with hypertension. Cochrane Database of Systematic Reviews
2005CD005182.
32. 32
40. Schroeder K, Fahey T, Ebrahim S. How can we improve adherence to blood pressure-
lowering medication in ambulatory care? Systematic review of randomized controlled
trials. Arch Intern Med. 2004;164:722-732.
41. Borzecki AM, Oliveria SA, Berlowitz DR. Barriers to hypertension control. Am Heart J.
2005;149:785-794.
Milliman Study
1) Kathryn Fitch, Kosuke Iwasaki and Bruce Pyenson. "Better Management Can Help Reduce the
Burden of Type 2 Diabetes: A 20-Year Actuarial Projection." Milliman, Inc. New York: March
2010. page 1, table 1.
(2) Ibid., page 3, line 56.
(3) Ibid., page 3, lines 5-9.
(4) Ibid., page 1, lines 29-34.
(5) Ibid., page 2, lines 40-1.
(6) Ibid., page 2, table 2.
(7) Ibid., page 15, table 10.
SOURCE Milliman Inc.
Captain James A Lovell
Federal Healthcare Center
Social Media Diabetes Research Study
David R Donohue M.A.
Research Project Coordinator