Transcript Project Title

Patient Safety/
Quality Improvement
Project Overview
Ariane Marie-Mitchell,
MD, PhD, MPH
Health Care Problem
• State reason for action
• Provide supporting data if available
Example: Health Care Problem
There is a much longer delay for 1st dose
delivery of antibiotics ordered between
11pm and 7am compared to day time
averages
This results in wasted time from the
nurses and pharmacists, wasted
medications, and poor quality of patient
care
3
Historical Data – The Problem
4
Stakeholder Analysis
• Who is involved? Who is affected?
• Try Mind Mapping
• Start interviewing- qualitative baseline data
Charge nurse
Nurses
Providers
Couriers
Patients
Patient
families
Pharmacists
Admin./PSR
Define the Aim
• What are you trying to accomplish?
• Specify
– numeric goals (how good?)
– time frame (by when?)
– patient population/system (for whom?)
K. Shannon, 2012
Aim for Quality Health Care
Safe — Avoid injuries to patients from the care that is intended to help them.
Safety must be at the forefront of patient care.
Timely — Reduce waiting for both patients and those who give care. Prompt
attention benefits both the patient and the caregiver.
Effective — Match care to science; avoid overuse of ineffective care and
underuse of effective care.
Efficient — Reduce waste. The health care system should constantly seek to
reduce the waste and the cost of supplies, equipment, space, capital, ideas, time
and opportunities.
Equitable — Close racial and ethnic gaps in health status. Race, ethnicity,
gender and income should not prevent anyone from receiving high-quality care.
Patient-Centered — Honor the individual and respect choice. Each
patient’s culture, social context and specific needs deserve respect, and the
patient should play an active role in making decisions about her ownIOM,
care.2001
Example. Specific Aim
• “Decrease delivery delay in 1st dose
antibiotic by 50% between 2300 and 0700
on units 6100-6300 by February 20, 2012”
Identify Measures of Change
• Outcome Measures
e.g. Duration of delay in administration of 1st dose
antibiotics
• Process Measures
e.g. % of antibiotic following forms filled
e.g. % of fax orders with telephone follow-up
• Balancing Measures
e.g. Staff satisfaction
10
Cause-Effect Analysis
• Fishbone Diagrams
• Flowcharts
Constructing a Fishbone
Cause and Effect Diagram
– Get the right people in the room
– State and clarify the “effect”
– Brainstorm list for 4 Ms/P involved in the
process or effect interested in
– Brainstorm causes for each of these
– For each cause ask “why” 5 times to get to
underlying causes
K. Shannon, 2012
K. Shannon, 2012
Example: Fishbone Diagram
Process
Communication
Unaware of effect on
Different levels of knowledge
Poor process understanding
Old fax machines
Unreliable tube system
Machines
Poor MD RN communication
Repeat orders via fax
Understaffed 2300-0700
ABX
Delay
No EMR = extra steps
# of pharmacy units open
System Failures
14
Constructing a Flowchart
Start/End
Start/End
Process/
Task
Input/
Output
Delay
Decisio
Decision
n
Sequence
Document
Prescription Renewal
Example:
Flowchart
The process begins when the patient requests renewal.
The Product of Value is that the prescription is delivered.
Call patient
(need more info)
Phone call
(enough info provided)
Name, DOB, Medication, Dose,
Frequency, Pharmacy
Review chart: medication sheet, last visit,
next visit, other parent or pharmacy info
(consistent)
(not consistent)
Discuss with clinician
(PA not needed)
(PA needed)
Call for Medicaid Prior Approval
Page 2
Select Change
• Focus on a change concept
eliminate waste, improve work flow, optimize
inventory, change work environment, improve
patient interface, manage time, reduce variation,
improve error proofing, improve service
• Perform effort vs yield analysis
High effort
Low yield
High effort
High yield
Low effort
Low yield
Low effort
High yield
Evidence Review
• Identify relevant literature or
best practice models
• Who else has thought about this
problem and tried to fix?
• Critically appraise and describe
how relates to your project aim
Example: Evidence Review
Background
• Pneumonia = 600, 000 Medicare hospitalizations/yr
• Previous Medicare Guidelines recommend antibiotic treatment within 8 hrs of
hospital arrival.
Methods
• Retrospective cohort study design
–
–
18, 209 Medicare patients (>65 yrs) hospitalized with community-acquired pneumonia
(July 1998-March 1999)
Outcomes:
1) severity-adjusted mortality (in hospital and 30 day)
2) readmission within 30 days of discharge
3) length of stay (LOS)
Conclusions
•
Antibiotic administration <4 hours of arrival was associated with
decreased mortality and LOS among a random sample of older
inpatients with community-acquired pneumonia who had not
received antibiotics as outpatients.
Houck, P, et al. Timing of Antibiotic Administration and Outcomes for Medicare Patients Hospitalized With
Community-Acquired Pneumonia. Arch Intern Med. 2004;164:637-644.
19
Test a Change
Repeat,
Tweak,
Stop
Analyze
Results
A P
S D
Act
Plan
Study
Do
Who,What,
When, Where
Collect
Data
Plan
• Specify a hypothesis
• Use SMART Objectives
Specific - have a single purpose
Measurable - tied to a result statement
Attainable/realistic - know your barriers and resources
Responsibilities clear – tie names to each objective
Time connected - clear completion dates
(consider creating a timeline)
Example: SMART Objectives
1. Interview charge nurse, nurses, couriers and pharmacists
during a full overnight shift (11p-7a) by day 7- Jose
2. Conduct a literature review of best practices by day 7Paymonh
3. Meet as a team to do cause-effect analysis by day 9- team
4. Collect quantitative data on initial state using Form 1a by
day 10- Brent and Craig
5. Meet as a team to discuss potential interventions by day 12team
6. Discuss intervention plan with pharmacists, couriers and
nurses and implement by day 16- Jose and Paymohn
7. Collect quantitative data on follow-up state using Forms 1b
and 2a by day 20 and analyze- Brent and Craig
22
Do
• Two types of data
1. Qualitative
e.g. interviews, focus groups, suggestion boxes
show quotations, summarize themes
2. Quantitative
e.g. anything that can be counted
show run charts, bar charts, pie charts…
• How did you collect the data?
• What sample did you use and what was your
reasoning?
Study
• Qualitative Data
• Quantitative
Data
– What did you
learn?
– Where does this
lead you?
Example: Qualitative Data
1. Parents observed speech problems by age 4 and as young as
11 months

“Around 10 months I noticed she was not starting to talk like my other
children did”
2. A minority of children received Early Intervention


“I asked the doctor about it but they were just like oh, he’ll grow out
of it”
“We were supposed to have a lady come to the house and stuff but it
was impossible because I was working full-time and he was at
daycare”
Example: Quantitative Data
100%
100% 100%
95%
93%
93%
87%
84%
80%
85%
91%
87%
85%
91%
100%
97%
93%
95%
90%
85%
75%
Percentage
Automated BMI% calculation
60%
55%
47%
Incorporated BMI%
into nursing data collection
All Practices
Our Data
40%
Goal (95%)
Added BMI% to vitals
21%
20%
0%
May
June
July
August
September
October
November
December
Janurary
February
March
Act
• How does your data inform your
understanding of the health care problem?
• How does your data influence your
interpretation of how to improve the system?
• What will you do next?
Sustainability
• What barriers do you perceive to sustaining
the change?
• What resources are available to sustain the
change?
• What is the cost-benefit of the old process
versus the new process?
References
• References for literature cited
• Names of faculty/staff interviewed
• Location of sites where processes observed, or
any other relevant info
≠
Quality
Improvement
Research
Aim(s)
Hypotheses
Next steps
Results
A P
S D
Act
Plan
Study
Do
Methods
Measures
Research
Quality Improvement
• Systematic investigation
designed to contribute to
generalizable knowledge
• Project involves introducing
an untested intervention
and data is being collected
to establish scientific
evidence of its efficacy
• Intervention is
demonstrated, known, or
widely accepted
• Project limited to
implementing a practice to
improve quality of care
• Performance data collected
for clinical, practical or
administrative purposes
only
IRB approval required
*** IRB approval not required
but ask if uncertain or if
considering publication
IHI Open School QI Practicum
• Optional
• Receive additional guidance and a Certificate
– Sign agreement with faculty mentor
– Submit aim/plan and receive feedback
– Submit cause-effect diagram, PDSA cycles, run
chart, and summary