POS lecture

1
POS lecture

• Quality improvement

John M A Bohnen Dec 2006
2
Overview

• Define quality
• Identify problems
• Discuss solutions

3

• degree to which health services for
  individuals and populations increase likelihood
  of desired health outcomes consistent with
  professional knowledge

Chassin et al JAMA 19982801000
4
What is quality of surgical care?

• high volumes?
• happy patients?
• good technique?

5
What is good quality?

• Alive 10 years after modified radical mastectomy
  or abdominoperineal resection

(clinical)
6
What is good quality?

• Alive 10 years after lumpectomy/rads or low ant
  resection, with intact breasts and no stoma

(clinical and functional)
7
?Broadening dimensions of quality

• Clinical
• Functional
• Cost
• Patient satisfaction

Dartmouth value compass 1990s
8
Dimensions of quality according to patient
outcomes

• safe
• effective
• patient centred
• timely
• efficient
• equitable

Institute of Medicine Crossing the quality
chasm. Nat Acad Press 2001, Washington
9
Dimensions of quality according to patient
outcomes

• safe
• effective
• patient centred
• timely
• efficient
• equitable

10
Harvard Medical Practice Study

• 51 hospitals, gt 30,000 discharges in 1984
• adverse events (AE) in 3.7,
• 58 of AE were errors, 28 negligence, 14 fatal

Brennan et al NEJM 1991 324370
11
Canadian Adverse Events Study

• 20 hospitals, 3,745 charts in 2000
• adverse events (AE) in 7.5
• 37 of AE were preventable, 21 fatal

Baker et al CMAJ 2004 1701678
12
Efficacy? effectiveness

• Effectiveness
• how something performs in real life
• Eg few patients comply because of cost, fatigue,
  impotence

• Efficacy
• how something performs in ideal circumstances
• Eg BP drug does great in randomized clinical trial

13
Why an efficacious treatment may be ineffective

• patient doesnt agree with treatment plan
• surgeon doesnt perform operation well

14
A short list for patient centred hospital care

• show respect and compassion
• find out what your patients want
• control their pain

15
(cont)

• plan discharges to avoid gaps in care
• try not to keep them waiting
• above all, inform and involve them

16
For patient populations, timely care requires
accessibility
17
Accessibility issues in Ontario

• Emergency departments
• intensive care beds
• hip, vascular surgery
• radiation therapy
• family doctors

18
Dimensions of quality according to patient
outcomes

• safe
• effective
• patient centred
• timely
• efficient
• equitable

19
We waste

• money
• patients time
• our time
• our energy

20
Costs include

• treatment
• treatment failure
• treatment complications
• risk reduction (eg public health)
• indirect costs (eg patients off work)

21
Who should get scarce resources?

• Should alcoholics get liver transplants?
• Should rich people be able to jump the queue?

Answers depend on our values
22
Two patients in ER

• Age 70, healthy
• calls in pain
• seen right away
• goes for Xrays, dx made in an hour
• gets analgesics right away, goes to OR sooner

• Age 70, dysphasic 2º to CVA
• agitated
• waits two hours for MD
• with translator, MD finds agitation due to pain
• finally gets abd XR free air

Is that fair?
23

• quality is multidimensional
• your values determine what you work on

Start measuring
24
Overview

• Define quality
• Identify problems
• Discuss solutions

25
How often is quality deficient?

• large U.S. study comparing actual care to
  recommendations
• Patients received 55 of recommended care
• worst was 11 (for alcohol dependence)
• best was 79 (for senile cataracts)
• study concluded that deficits in adherence to
  recommended processes pose serious health threats

McGlynn et al NEJM 20033482635-45
26
Lets put these issues into a framework

• Start thinking about system problems, not bad
  doctors

27
System

• An interdependent group of items, people, or
  processes with a common purpose

Langley et al The improvement guide Jossey-Bass,
San Francisco, 1996
28
Donabedian 3 targets to assess health care
systems

• Structure
• Process
• Outcome

Berwick et al Curing Health Care, Jossey-Bass,
San Francisco, 1990
29
Structure

• The scene before a process begins

30
Examples of structures

• hospital equipment
• number of doctors and nurses
• number of training positions
• rules, legislation, hospital bylaws

31
Process

• A series of steps that leads to an outcome

32
Example of a processes administering a drug

• Pharmacy buys and stores drug? doctor orders it ?
  pharmacist provides it? nurse administers it

Each step is a mini-process with smaller steps
33
Structure and process are similar but different

• Similar
• both lead to outcomes
• developing a structure is a process
• Different
• structure is more static
• structure is measured more easily (less a moving
  target)

34
Structures and processes can be described and
measured

• By counts (eg number of nurses or operations)
• By map, diagram or pictures (eg practice atlas)
• By narrative
• etc

Describing a process usually helps us to
understand it
35
Managing severe abdominal pain
Establish diagnosis
no
Diffuse peritonitis?
Abdominal pain
yes
normal
amylase
operate
hi
Treat pancreatitis
36
Good quality desired outcomes

• What are those outcomes?
• They are multidimensional (SEPTEE)

providers of care must attend to many potential
outcomes
37
What problems in structures and processes lead to
unwanted outcomes?
38
Chassin et al have classified problems with
quality

• underuse
• overuse
• misuse

JAMA 19982801000
39
Example of Underuse

• preop antibiotic prophylaxis

40
Example of overuse

• prolonged prophylactic antibiotics for joint
  replacement

41
Example of misuse

• too early preop antibiotic prophylaxis

42
We face many barriers to high quality care, such
as

• inadequate human and capital resources
• user unfriendly processes
• cannot keep up with changes
• cost
• human nature
• geography (eg remote, rural)

43
To understand barriers, lets revisit the health
care system

• System An interdependent group of items,
  people, or processes with a common purpose
• Do providers and consumers have a common purpose?

If not, then health care is not a single system
44
We work in a matrix of care who are the players?

• patients and families
• providers

45
Other players

• payers
• trainees
• regulators
• legislators
• media
• courts

46
Who else?

• Community services
• hospitals
• regional health authorities
• numerous health professions
• etc

We dont always work together, and sometimes we
compete, because our needs and tasks differ
47
We are governed especially by

• CPSO
• Hospital governance
• Work contract
• RCPSC

48
We have disparate goals, eg

• Patients get and stay well
• Providers get paid, help patients
• Payers control costs
• Hospitals control costs, ace report card
• Trainees learn craft, get jobs, control debt
• Legislators enhance care, get votes
• Regulators manage risks, protect public
• Media sell copy, sell ads
• Courts find fault, provide redress

49
Even players on the same side face conflicting
demands, eg

• Information security impedes access to
  information needed for patient care
• Increased accessibility adds costs

50
How to improve?

• Decide what to improve (aim)
• Find out whats going on (measure)
• Plan how to improve on it
• Implement the improvement
• Measure to see if the improvement worked
• If not, try another approach
• If so, hold and increase the gains

51
Deming

• developed statistical process control into a
  philosophy and science
• revolutionized postwar Japanese industry
• is considered the prophet of quality improvement

52
Deming
Problems (and solutions) of quality are usually
built into complex production processes, not lack
of will or skill of the workers
53
The improvement project
54
What are we trying to accomplish?
How will we know that a change is an improvement?
Plan
Act
What changes can result in improvement?
Study
Do
55
Aim
Measures
Plan
Act
Tests of change
Study
Do
56
Starting

• Gather a team
• Include all involved

57
Aim

• should be value-driven (eg improve patient
  centredness)
• should inform improvement
• should engage caregivers enthusiasm

58
How should we measure quality?

• What is the purpose?
• What should we measure?
• Who should do it?
• How should they do it?
• Where should the results go?
• How should we use the results?

59
What to measure?

• Which quality dimensions?
• consider needs (relevance) and resources
  (practicality)
• structure, process or outcome?
• structure/process may be valid surrogate for
  outcome

If linked closely with outcomes
60
Who should do it?

• Decide whether clinicians, bean counters or both
• Need broad input to prevent failure or irrelevance

61
How should they do it?

• count on fingers
• pen and paper
• spreadsheet
• fancy stuff
• More fancy generates ? technology support is it
  worth it?

62
Where to send the results?

• Who should know, for improvement?
• Who is accountable? (whos the boss, whos to
  blame)
• Who generated the data?
• caregivers
• patients
• Whos paying?

63
Should we sent to

• media
• courts
• consumers groups
• CPSO
• RCPSC
• our parents, grade school teachers

64
Dissemination of results may encounter resistance

• dont underestimate the potential for problems
• be creative and sensitive
• involve people in advance and warn the rest

65
What to do with the data

• common sense, leadership, culture
• will tell you what to do
• those that start measuring often end up leading
• sometimes measurement drives change

66
Start measuring

• problems and solutions may become obvious
• we like to see our results
• we naturally want to improve
• basis for formal improvement work

67
Fixes

• Measure and report quality routinely in large and
  small scale (eg province and clinical service)
• Use information technology better
• Educate and inform patients better
• Implement effective financial incentives

Steinberg Improving the quality of care can we
practice what we preach? NEJM 20033482681
68
Other fixes

• Apply volume standards (Birkmeyer et al)
• Leadership and culture (Inst for Healthcare
  Improvement - IHI

69
Volume a outcome?

• Birkmeyer et al for (1)
• VA studies against (2)

• Birkmeyer et al Surgery 2004135569
• Khuri et al Ann Surg 1999230414

70
Financial incentives

• Epstein et al Paying physicians for high-quality
  care. NEJM 2004 350406

71
Computer-based decision support(CDSS)

• Hunt et al Effects of computer-based clinical
  decision support systems on physician performance
  and patient outcomes. A systematic review JAMA
  19982801339 ?were getting there

72
Success stories

• CABG mortality (1)
• Antibiotic decision support (2)
• VA/ACS (3)

OConnor et al JAMA 1996275 841 2) Evans et al
NEJM 1998338232 3) Khuri et al Arch Surg
200213720
73
Best evidence
Effective management practice
Decision Support
Local measurement