B E T H T A Y L O R , D C N , R D , F A N D
HOW TO CONDUCT A RESEARCH/PI/QI PROJECT
OBJECTIVES
• Understand common misconceptions about clinical
research
• Differentiate between projects that need
Investigational Review Board approval and those
that do not.
• Describe common obstacles and how to avoid
them.
DEFINITIONS
• Research • Outcomes based – “the study of the end results of health
services that takes patients’ experiences, preferences, and values into account—is intended to provide scientific evidence relating to decisions made by all who participate in health care.”1
• Patient Centered – “research that addresses the questions and concerns most relevant to patients, and ...” www.pcori.org
• Quality Improvement – focuses on process improvement2
• Performance Improvement – focuses on improving institutional and individual performance2
WHY SHOULD NUTRITION SUPPORT CLINICIANS DO NUTRITION
RESEARCH/PI/QI?
• Improve practices in order to improve patient
outcomes, reduce cost, etc.
• Improve skill set.
• Increase our value to the healthcare team.
• If we are left out, others won’t do it right.
WHY DON’T WE DO MORE?
• Not enough time
• No funding
• Lack of leadership support
• Lack of administrative support
• Knowledge deficits:
• Statistics
• Research methodology
• IRB process
JAND epub 2015: Why RDs are not doing research
RESEARCH VS. PI/QI
• To test a hypothesis or establish practice standards
• May put patients at risk
• Answers a research question.
• To assess or improve a process, program, or system or improve performance.
• Does not put patients at risk
• Improves or creates a process, program or system that increases safety, efficiency or satisfaction.
Modified from AND webinar by DellaValle DM 2015
RANDOMIZED TRIALS
• Assemble a group of individuals, randomly
assign them to intervention groups, follow them
forward for a period of time for outcome(s)
Control group
Disease No Disease
Treatment group
Disease No Disease
Group of study participants
Randomiz
e
RANDOMIZATION
• Goal of randomization:
• Make the intervention groups identical with respect to all
factors except for the intervention of interest
• Remove confounding factors
• Like “tossing a coin” to decide which group a study participant is assigned to.
• More labor intensive type of research.
COHORT STUDIES
• Assemble a group of individuals (the cohort),
collect data on exposure(s) and subsequent
outcome(s)
Unexposed
Disease No Disease
Exposed
Disease No Disease
Cohort of individuals
STEPS IN THE PI/QI OR RESEARCH PROJECT
• What outcomes to track/develop study question
• Literature review
• Brainstorming with key players
• Project Plan/Study protocol
• Approval from colleagues, MDs, involved depts
• Investigational Review Board (IRB), Human Subjects
Committee ?
• Implementation
• Data analysis
• Sharing the findings
IDEA FOR THE PROJECT – FINDING YOUR INSPIRATION
• Identify problems in systems, practices, equipment,
etc.
• Identify potential outdated systems, practices,
equipment, etc.
• Difference of opinion regarding best practice
• Find something you are excited about
• Look through poster abstracts from CNW, FNCE
• Pair and share
EXAMPLES OF OUR INSPIRATIONS
• A seemingly large number of inappropriate TPNs
• Prolonged periods of NPO to oral diet and/or no tube
feeding peri-op
• Difference of opinion of the nutrition support RDs regarding
the need to check serum triglycerides in PN patients
WHAT OUTCOMES WILL YOU TRACK ON WHICH PTS/WHAT IS YOUR STUDY
QUESTION?
• Don’t try to do too much at once.
• Narrow down your topic
• If you are doing research, use PICO
• If you are doing PI/QI, you can do a modification of
PICO to narrow your focus.
• Remember: If you cannot measure it, you cannot
improve it. William Edwards Deming
QUESTION/HYPOTHESIS
• PICO helps to structure the research idea (for an
observational or intervention study) into a
detailed clinical study question or hypothesis and
can be used for the literature review:
• Population of interest
• Intervention to be tested
• Comparison strategy
• Outcome(s)
EXAMPLE OF PICO
• P – In adult ICU trauma patients on tube feeding
• I – immune enhancing formula
• C- standard formula
• O- LOS, time on vent, infections, etc.
• Among adult ICU trauma patients, does the use of
an immune enhancing formula as compared to an
isocaloric, isonitrogenous standard formula have a
statistically significant effect on LOS, time on vent,
infections, etc?
PI/QI PROJECT IDEA
• P- in patients on tube feeding
• I - ready-to-hang formula delivery
• C- cans
• O- % of goal delivered, nursing satisfaction, waste of
tube feeding.
• Pair and share
OTHER IDEAS
• Percentage of patients discharged home on tube
feeding who are readmitted within 30 days with
nutrition-related problems.
• What percentage of malnourished patients are and
are not identified with your facility’s malnutrition
screening tool?
LITERATURE REVIEW
• Has your question already been answered?
Has your project already been done?
• Critically review the literature – look for
differences in methodologies.
• Maybe all you need is a literature review to
convince others to change practice.
• Get a librarian’s help, if available.
• Also check the Academy’s Evidence
Analysis Library
DOING YOUR SEARCH
Get search terms from your PICO
MeSH terms, keywords
Boolean terms OR, AND, NOT
Consider whether you want your search to be wide
or narrow. Usually start with wide.
Learn how to use Ovid, save searches
In PubMed, sign up for an account, save searches,
have search updates sent to you.
Check references from the articles you find
Click on “related articles”.
Save articles in Endnote or Refworks.
EXAMPLE OF A SEARCH
• Using our PICO example
• MeSH terms for P – adult;patients; wounds and
injuries
• MeSH terms for I – enteral nutrition;
• MeSH terms for O – length of stay;
respiration/artificial; infection
• Also look for relevant keywords – trauma; tube
feeding; mechanical ventilation; immune-
enhancing (no MeSH term for this), etc.
CRITIQUING THE LITERATURE
• To stay organized, keep track of info as you read
the studies
• Create a spreadsheet
• Patient Population, Sample size, Intervention,
Comparison, Results, Limitations
IT TAKES A VILLAGE
• Make it easier on yourself and increase your
chances of doing it right the first time.
• Don’t try to do it alone.
• Early in the process -Who can help you?
• Share your idea with colleagues. Maybe someone
has a similar idea. Maybe someone is already
doing the same thing.
• Get your manager’s support
• Internship at your facility? Suggest your project or
part of your project as an intern project.
YOUR TEAM
• Get input from those with more experience doing
research and PI projects.
• If you are new to this, ask for a mentor – MD, basic
scientist, another RD
• Identify team players, stakeholders, etc.
WHO TO INCLUDE?
• Pair and share
• Give examples of past projects and who you
collaborated with.
• Who would you collaborate with on a future
project?
BRAINSTORM
• Let the creative juices flow with “your team”.
• Work on the protocol with the team.
STATISTICIAN/RESEARCHER
• Get the help of a statistician, if possible.
• Better to know early how to record your data to be
most useful during the analysis phase, what sample
size you need (if it is more than a pilot study), etc.
• Bias
• Statistical inference/Sample size/Power
• Generalizability
DIETETICS PRACTICE BASED RESEARCH NETWORK
• Academy of Nutrition and Dietetics – Practice
Based Research Network – free to members
• MyAcademy profile, New Enrollment, and click on
this group.
• http://eatright.org/members/DPBRN/
• Webinar
• Research fact sheets
PROTOCOL
• Rationale
• Objective
• Methods
• Populations
• Time frame
• Expected outcomes
PROTOCOL DEVELOPMENT
• What type of study are you going to conduct?
• Observational? Interventional? Survey?
• If interventional: What is your “exposure” of interest? How will you measure it?
• What is your outcome of interest? How will you measure it?
• What is your study population?
• How will you collect the data for your study?
• What are methodological issues that you need to consider?
APPROVAL
• Before finalizing proposal, present proposal to
relevant departments (RDs, nursing, PCTs, MDs)
• for practical advice and approval of procedures that impact patient management
• suggestions about educating staff, flyers, reminders
IRB? PI/QI VS RESEARCH?
• Check with your institution’s IRB.
• Exempt (PI/QI), expedited or full committee
• Expedited or Full Committee if:
• Your intervention is not standard practice
• Randomizing
• Subjects’ anonymity can’t be guaranteed
• Risk
• Other institutions are involved.
• Results are going to be used for generalized knowledge.
QI
• http://www.hrsa.gov/quality/toolbox/methodology
/qualityimprovement/part3.html
• This site is a wealth of information about QI – how to
conduct, forming teams, etc.
IRB APPROVAL PROCESS
• Major considerations for IRB committee
• Risks to participants vs. benefits for participants and
society
• Consent process (should present risk/benefit information
adequately and fairly and minimize potential for coercion)
• All investigators must provide certificate of
Education on the Protection of Human Subjects
(CITI training) before beginning research.
• www.citiprogram.org
IMPLEMENTATION
• Follow the time frame that you set in your protocol.
• Use whatever method works for you to follow that
time frame – paper calendar, electronic reminders.
• Set up regular meetings with your “team”.
• Do a pilot of data collection to test your data
collection form. Modify it, if need be.
SHARE YOUR RESULTS
• Don’t be selfish with your results. Don’t file them away yet. Share them.
• Ask those who helped what they think the implications of your results are. What should be done next?
• Take results to relevant committees for their input.
• Submit a poster of your work to a conference– look for opportunities to display your poster at your facility.
• Write an article.
• Consider who you want your audience to be.
TAKE-HOMES
• Find your inspiration
• Do literature review
• Share your idea with others
• Enlist help of others
• Make a detailed plan – study question, PICO,
outcomes to track, data collection form
• Implement
• Share your results
REFERENCES
1. Clancy CM, Eisenberg JM. Outcomes research measuring the end results of health care. Science. 1998; 282: 245–246.
2. Bornstein, T. Quality Improvement and Performance Improvement: Different Means to the Same End. http://fkilp.iimb.ernet.in/pdf/Healthcare_Quality/Concepts
3. JAND epub 2015: Why RDs are not doing research
4. AND Webinar by DellaValle DM. Registered dietitian/nutritionist guidelines for successful quality improvement implementation: Getting started