Renal Mra Template - PowerPoint

Document Sample
Renal Mra Template - PowerPoint Powered By Docstoc
					 Clinical Research for
Community Hospitalists

     Peter Lindenauer MD MSc
      Baystate Medical Center

     Lakshmi Halasyamani, MD
     St. Joseph Mercy Hospital

    Andrew Auerbach MD MPH
University of California San Francisco
• What is research?
  • Is there good research? Bad research?
  • Why should I spend my time doing it?
  • Questions, questions everywhere
• Components of a research project
  • A general outline of what you should be
•       Research project case studies
    •     Teaching points
         1.   Designing the project
         2.   Working with the IRB
         3.   Project Funding
         4.   Data issues
         5.   Authoring and presenting a project
         6.   Challenges and opportunities
•       A real-life research success story
         What is „Research?‟
• Not just GUSTO RCT‟s:
  • Training (or mentorship) heavy:
     • Meta-analyses
     • Decision analysis
  • Training (or mentorship) „lite‟:
     • Surveys
     • Quality improvement
  • Intermediate requirements
     • “Outcomes” research
          Why do Research?
• General reasons:
  • Expand the body of medical knowledge
  • Help your patients
  • Establishing a professional niche
• Personal Reasons
  • Professional satisfaction
  • Fun
  • Fame and fortune
• Specific (or local) reasons
  • Answer a question important for your institution
  • Answer a question important for your practice
    The “Research Question”
   “The uncertainty about something in the
   population that the investigator wants to
   resolve by making measurements on his/her
   study population”
• AKA: What you would like to know
• There are no shortage of these
   Initial Research Challenges
• Getting over the fact that you are not a
  • Get over it
• Finding important questions
• Transforming these into study plans that
  • Feasible
  • Valid
   Questions ARE Everywhere
• Build on experience
  • Mentorship
  • Scholarship
• Be alert to new ideas and practices
  •   Skeptical attitude about prevailing beliefs
  •   New technologies
  •   Observation of patients, new literature
  •   The need to improve – ie quality improvement
• Develop questions in an iterative process
  • Consultations with advisors, colleagues
  • Creativity and Tenacity
 Step 1 – Refining the Question
• Research questions often start out vague
  • Ex: Are fewer nurses bad for patients?
• To carry out a study you need to be able
  to measure things
• This means you need to get specific
    Ex: For medical patients hospitalized in a non-
    ICU setting is there an association between
    the patient to nurse ratio and in-hospital
Sample Research Question - 1
  Are catheters harmful to patients?

  What‟s wrong with this research question?

  How might you improve it?
Sample Research Question - 2
 Does phosphorous supplementation reduce in-
    hospital mortality in patients with acute
           exacerbation of asthma?

What problem might you have in trying to carry
out this study?

What other outcomes might you focus on?
Sample Research Question - 3
Do electronic reminders to physicians caring for
  patients with pneumonia increase in-hospital
        pneumococcal vaccination rates?

What are the strengths of this research

Do you think the results of this study would be
widely generalizable?
      Planning your research
• What do you think is going to happen (or
  has happened)?
  • AKA – The Hypothesis

  “Implementing an enoxaparin guideline will increase
             appropriate use of enoxaparin”
 Characteristics of Good Research
• Feasible
   • Adequate number of subjects
   • Adequate technical expertise
   • Affordable in time and money
• Interesting
• Novel
   • Confirms, extends or refutes previous findings
• Ethical
• Relevant
   • To scientific knowledge
   • To clinical and health policy
   • To future research directions
       Planning your research
• What do you want your project to do?
  • AKA – The Specific Aims

 “We will measure use of enoxaparin before and after our
                guideline is implemented”
       Planning your research
• How are you going to measure the effect
  of the intervention?
  • AKA – Your Outcome

“We will count the number of eligible patients who receive
  enoxaparin and calculate the percentage in each group
                that adhered to the guideline”
         Planning your research
• How are you going to be sure that other factors
  don‟t account for your findings?
   • AKA – Confounders or independent variables

“We will also look for whether patients had an allergy to enoxaparin, or
            whether they had a contraindication to enoxaparin”

   Unless you are doing a randomized trial, you should be thinking about
                          collecting this information
        Planning your research
• Where will the data come from?
  • AKA – Data Sources

 “Our quality-improvement nurse will abstract charts of 70
   eligible patients cared for before and after the guideline
                        was put in place”
        Secondary Data Sources
• Why Secondary data?
   • How much time would you like to spend in medical records?

• Hospital discharge database
   • Eg: UB-92 or Discharge abstract
   • Contains information about admit, DC, Dxs, Procedures, MDs
   • Example : Who were the patients discharged during the month of
     July who were treated by a particular physician

• Cost accounting systems aka Financial Decision Support
   •   Eg TSI, HMC, McKesson
   •   Detailed utilization data
   •   Costs and charges
   •   Do not contain results
   •   Example: How often do we order an MRI‟s on stroke patients?
 More Secondary Data Sources
• Departmental level databases
   • Eg Radiology database
   • Strengths : Sometimes has more clinical information than
     contained in other db
   • Weaknesses: Usually lacking other important data
   • Example: Who were the patients that underwent a VQ scan
     during the month of July

• Data warehouse
   • A generic term to describe a database housing data for analytic
     rather than transactional purposes, often from multiple data
   • Look for these first
  Working with Secondary Data
• Find a research question to fit an existing
   • You learned that the Case Managers at your hospital
     routinely collect information about whether patients
     are at appropriate “level of care”
   • You decide to study whether Hospitalists are more or
     less likely to have non-qualified days
• Find a database to fit an existing research
   • You are interested in learning whether Hospitalists at
     your institution provide better care for patients with
     AMI than non-Hospitalists
   • JCAHO Core Measures database
       Planning your research
• How will you analyze the data?
  • AKA – Your analytic plan

    “We will do a side-by-side comparison of patient
   characteristics and percent adherence in each group”

              “We will hire a statistician”
         Lies, damned lies, and statistics

•       Don‟t let it be a deal breaker
    •     Finding a statistician or epidemiologist will follow
          your search for mentorship.
    •     If you consider the clinical question, statistics will
          •   Pay attention to confounders and outcomes when you think
              about your study.
      Lies, damned lies, and statistics

•    You can do some preliminary work yourself:

    1. General information web pages

    2. Online power/sample size calculators
 Why SIZE sometimes does matter
• There are various recipes for calculating sample size
   • Example - Comparing proportions of dichotomous variables
     using the Z statistic
• Components
   • Proportion of subjects in group 1
   • Proportion of subjects in group 2
   • Alpha
       • Statistical significance; one or two tailed
       • The probability of rejecting the null hypothesis if it is true
   • Beta
       • 1-Beta = Power
       • The probability of failing to reject the null hypothesis when it is false
• Formula
   • Look it up. Better yet, don‟t. Use a table or the web
       A typical study QI study
• You wish to compare the proportion of patients who
  receive DVT prophylaxis before and after your guideline
• The baseline rate is 30%
• You expect to increase the rate to 50%
• You set your alpha at .05, Power = 80%
• How many patients will you need to study?
 DVT Prophylaxis Intervention
• When
  •   Alpha = .05 and is two-tailed
  •   Beta = .20 ; ie Power = 80%
  •   p1 = .3, p2 = .5
  •   n = 93 in each group or 186 total
            A Stroke Intervention
• You hypothesize that better glycemic control may
  improve the outcomes of diabetic patients with ischemic
• You believe this intervention may reduce in hospital
  mortality rates by as much as 20%
• The mortality from stroke at your hospital is 10%
• How many patients will you need to enroll to prove your
   •   Alpha = .05 ; Beta = .2
   •   p1 = .10
   •   P2 = .08
   •   N= 50? 100? 250? 500? 1000? 5000?
             A Stroke Intervention
• You hypothesize that better glycemic control may
  improve the outcomes of diabetic patients with ischemic
• You believe this intervention may reduce in hospital
  mortality rates by as much as 20%
• The mortality from stroke at your hospital is 10%
• How many patients will you need to enroll to prove your
   •   Alpha = .05 ; Beta = .2
   •   p1 = .10
   •   P2 = .08
   •   N= 6418
           Take home points
• Its better to think about sample size before
  you begin your study
• Its not as hard as it seems
• Have someone double check your
  estimates and your calculations
• Its still a good idea to involve a statistician
        Planning your research
• Do I need to get my research approved
  through an IRB?
  • The answer is almost always YES

If your study will touch patients or their records it is always
      safer to approach the IRB than to proceed without
                 Ethics and the IRB

•       Always send your protocol to an IRB
    •     Every hospital that permits research is
          supposed to have one
    •     Find it, use it even if you think your research
          poses no risk at all
    •     Only situation where you may not need to
          consult the IRB – QI work that you will not
Ethics – not just for accountants anymore

•   Ethical conduct: an absolute requirement
    1. General research bioethics
    2. Human subject protection training
    3. Consent
           Lies, damned lies, and HIPAA

•    Pay attention to HIPAA when collecting data

                           Whatever that is….

    General rules:
       •    Pay attention to your institution‟s HIPAA rules
       •    Don‟t collect patient identifiers at all, if you don‟t need to
       •    Try to replace or remove patient identifiers wherever you can
       •    Don‟t send anything that could potentially be linked to
            patients over the internet or by mail
      Planning your research
• Do I need MONEY?
  • Your answer is almost always YES
  • The reality is usually different.

• Do I need HELP?
  • Know thyself
  • Seek mentorship, reality checks
         Research on $5 a day
• Internal projects
  •   Students and residents
  •   Department of nursing, nursing students
  •   QI department
  •   Mentors
• External projects
  • Secondary data analysis
  • Multihospital collaboratives
  • Statewide or national QI or safety initiatives
    Data Collection and Management

•   Develop a data abstraction tool
•   Ask other people to look at your tool
•   Pilot the tool on 10 charts
•   Fix the tool
•   Use software to facilitate data entry
    • Eg Teleform

•       Where to get it:
    1. Local medical school or school of public health
          •   Don‟t be afraid to talk to a lot of people
    2. Faraway medical school or school of public health
          •   Don‟t overestimate usefulness of the Internet or phone
    3. SHM
          •   Clinical ListServe, research committee
          •   Meet the professor session
•       Where NOT to get it
    •     Pharma
Planning your research

 Write it all down someplace.
           On paper.

         No, really.
         Outline of a Study Proposal
         ResearchSpeak                    English Translation

Research Question and         What questions will the study address?
Hypothesis                    What do you expect to happen?

Background and Significance   Why are these questions important?
                              What do we know about them?

Aims                          What do you plan to do?

Subjects                      Who will you study?
Inclusion/Exclusion           How will they agree to participate?
Design                        How will you actually do the study?
                              What is your „protocol‟?
        Outline of a Study Proposal
        ResearchSpeak               English Translation
Data Sources            Where will you get your data?
                        What do you plan to collect?

Outcome variables       Which data is key to your question (or

Predictor variable      Which will confuse (or confound) the

Statistical Issues      How large is the study and how will it be
Human subjects          How will you maintain ethical standards?
                        How will you protect patient‟s rights?
    Research Case Vignette 1
• Clinical Scenario:

        Your institution will be reviewed by the
        JCAHO for its use of patient restraints.

     The hospital feels it needs to improve restraint
        documentation and comes to you for help.
    Research Case Vignette 1
What might you do to address this problem?

Name a research questions that might arise
 from your intervention?
     Research Case Vignette 2
• Clinical scenario

One of the large insurance plans in your area has identified
             pneumonia as a health care priority.

  You are given a clinical pathway which the insurance
  company believes can improve the time to first antibiotic
           received and reduce hospital costs.
    Research Case Vignette 2
What would be the key outcomes in this
How would you design this study?
Where might you obtain data about cost?
Where might you obtain data about clinical
     Research Case Vignette 3
• Clinical scenario:

  You care for many patients with stroke and have
  developed a healthy skepticism about the battery of tests
  (Echo, MRI/MRA, TCD) recommended by the
  neurologists at your hospital
  Research Case Vignette 3
How might you translate this skepticism
into a feasible research question?

What kinds of outcomes would you focus

How might you obtain data for this study?
    Research Case Vignette 4
• Clinical Scenario:

 You recently learned that pneumococcal and influenza
 vaccines are underused by physicians – both in your
 local group and across the state

 You are curious why physicians are not adhering to
 national recommendations
  Research Case Vignette 4
What kind of study might you think of to
help answer the question

What are the challenges of carrying out
       A Research Success Story

• Background
  • Saint Joseph Mercy Hospital is a 478-bed community
    teaching hospital located in Ann Arbor Michigan
  • There are approximately 10,000 internal medicine
    admissions per year
  • The hospital was interested in improving the care of
    patients with heart failure and specifically improving
    the 30-day readmission rate
            The Research Checklist

•   Designing the project
•   Working with the IRB
•   Project Funding
•   Data Issues
•   Authoring and Presentation
•   Challenges and Opportunities
 Designing the Project: Heart Failure at

• 1998- 2000: 1406 admissions with a D/C
  diagnosis of HF (discharged in DRG 127)
• 227/1406 admissions were readmissions within
  30-days of discharge
• 30-day readmission rate was 17%
• National benchmarks for 30-day readmission
  rates are 6-8%
 Designing the Project: Readmissions at

• 50% of the readmissions were for cardiac
  reasons such as:
  • Heart Failure           -- Atrial fibrillation
  • Myocardial Infarction
• 50% of the readmissions were for non-cardiac
  reasons such as:
  • Acute renal failure     -- Urosepsis
  • Pneumonia               -- Stroke
  Designing the Project: Team-Building

• Enlist the support of key decision-makers
  at your institution
• Focus on a multi-disciplinary approach
• Align primary project goals with goals of
  the institution
  Designing the Project: the Research

• Does a comprehensive inpatient/outpatient
  heart failure program decrease the
  hospital‟s 30-day readmission rate for
  patients with heart failure?
  Designing the Project: Developing the

• Multi-disciplinary approach (physicians, nurses,
• Take the show on the road: present the
  intervention plan to all of the potential users
• Generate buy-in for the intervention: research
  meets reality
       Designing the Project: The Inpatient

• Heart Failure nurse reviews patient‟s care regarding:
   •   ACEI use
   •   Beta-blocker use
   •   Vaccination status re: pneumovax and influenza
   •   Need for indwelling foley catheter
   •   Safety with ambulation and need for PT
• Patient/Family Education
   • Fluid management                          -- symptom
   • Medication Adherence              -- sodium restriction
   • weight monitoring                 -- follow-up
   Designing the project: the Follow-up

• Intensive Intervention: Heart Failure Clinic

• Intermediate Intervention: Telescale Program

• Lower Intervention: Usual Care
        SJMH Heart Failure Intervention

                           Heart Failure Program

                            Patient identified with HF
                      (IV diuretic and clinical assessment)

                           Inpatient HF nurse intervention
                                    and Follow up

Lower Intervention               Telemanagement          Intensive Management
 Patients is managed by         NP management with            Patient is referred
  M.D. through regularly       Cardiocom® software and          to CHF clinic
scheduled appointments            use of Telescale®.
             Working with the IRB

• If goal is to publish the results, then seek IRB
• Identify contact person in institution‟s research
• Review application and committee meeting
• Submit proposal to the IRB
               Project Funding

• Sources within the Institution
• External Funding Opportunities

• Align Project Design and Development with
  Funding and Resources Available: Research
  meets Reality
                  Data Issues

• Data Collection
  • Content: outcome variables of interest
  • Method: electronic retrieval whenever possible
  • Sources: Cost accounting department, hospital
    information system, social security death index
• Data Analysis
  • Statistical Support
  • Power Calculations
          Current Status of Project

• Randomized Control Trial of HF program
  compared with usual care
• 1/3 of patients have been enrolled
• Anticipate study completion in one year with
  publishable study results
     Challenges and Opportunities

• Challenges
  • development of multi-disciplinary team
  • obtaining project funding
  • statistical support content expert may not be within
• Opportunities
  • contribute to the literature in a meaningful way
  • community settings are common places that large
    amounts of healthcare are delivered ie.
  • increase the awareness of the “system”
    contributions that Hospitalists can make
       Training opportunities
• Summer courses:
  • Harvard Clinical Effectiveness Course
  • Johns Hopkins Clinical Epidemiology training
    • 7-8 weeks in summer
    • Option to return following summers for continued
    • Johns Hopkins online training

Shared By:
Description: Renal Mra Template document sample