Quality Improvement

1
Quality Improvement

• Eric Mortensen MD, MSc, FACP
• VERDICT Research Program

2
Goals

• To understand why there is currently an emphasis
  on improving the quality of health care
• What are PDSA cycles
• What are the barriers to successfully assessing
  quality of care

3
Introduction

• Improving the quality of health care is currently
  a hot topic
• Agency for Health Care Research and Quality
• Institute of Medicine - To err is human
• National Quality Forum
• Why are high error rates tolerated in health care
  but not in other industries?

4
Reasons for Interest in Improving the Quality of
Health Care

• It is the right thing to do!
• High costs of health care
• Can we reduce costs if we reduce errors?
• Unexplainable variations in practice
• Similar patient treated differently
• Underutilization of services
• Overutilization of services

5
Costs of Health Care

• Medical care was 6 of GNP in 1980 and 15.2 in
  2003
• Employers are increasingly unwilling to cover
  these costs
• Although expenditures stabilized in the late
  1990s we are now experiencing 5-10 increases in
  health costs each year

6
Reasons for Increasing Costs

• Rising intensity of services
• ICU
• Transplants
• Medications
• Increasing life expectancy
• Overuse and under-use of medical technology

7
Practice Variation

• Variations in practices have been documented in
  many areas of medicine
• Four-fold difference in radical prostatectomy in
  adjacent counties
• Variations in length of stay of pneumonia,
  congestive heart failure, and myocardial
  infarctions
• Rate of use of breast conserving surgery

8
Medical Errors

• There are an estimated 44-98,000 patients killed
  each year by medical errors
• Errors difficult to define
• Easy
• Wrong medication or dose
• Operating on wrong leg
• Hard
• Complications of needed therapy
• Anaphylactic reaction to medication

9
Comparison to Other Industries

• Firestone recall
• 6,500,000 tires involved
• 119 deaths attributed to defects
• 0.0000183 deaths/tire
• Medicine
• 33,600,000 admissions per year
• 44,000 deaths
• 0.0013 deaths/admission

10
What Do You Need to Assess Quality?

• Valid outcomes
• Risk adjustment method
• Reliable data

11
Defining Outcomes

• Evaluating quality assumes that one can define
• The correct set of things that should happen
• What is the expected outcome?
• The group of patients associated with an outcome
• Are patients with diabetes expected to have the
  same complication rates as those without diabetes?

12
Outcomes Measures

• Some outcome measures are obvious
• Mortality during hospitalization
• Complications after surgery
• Problems with these outcomes
• What is the correct complication rate?
• What is the expected mortality rate?

13
Types of Outcomes

• Processes of care
• of patients with HTN with BP lt 140/90
• Initial antibiotics within 4 hours for pneumonia
  patients
• Intermediate outcomes
• CAD patients with elevated LDL brought under
  control
• True outcomes
• Mortality
• Incidence of ESRD in diabetics

14
What is a Medical Error?

• Anaphylaxis due to previously unexposed
  medication
• Ototoxicity due to high dose gentamicin therapy
• MD not seeing test result until next day
• Wrong dose of medication in patient with CRI
• Surgery on wrong leg

15
What is Needed to Have a Good Definable Outcome?

• Easily measured
• Data available
• Clear definition
• Information reliably obtained
• Scientifically valid
• Strongly associated with outcomes
• Mortality
• Patient satisfaction

16
Example of Defining Quality Outcomes

• National panel of cardiothoracic surgeons asked
  to determine what is the appropriate failure rate
  for artificial valves in five years

17
Number of CT Surgeons Opinions on the of Valves
that will Fail
18
What Do You Need to Assess Quality?

• Valid Outcomes
• Risk adjustment method
• Reliable data

19
Data

• Reliable data is often difficult to find
• May need to gather it yourself
• If available you need to ask why it was gathered
  and if this may bias it
• Use of administrative databases
• Example
• Discharge diagnosis of UTI versus MI

20
Risk Adjustment

• Uses statistical techniques to adjust for
  severity of illness
• If one physician only operates on easy cases and
  another only works on difficult cases is it
  really fair to compare the two?
• These techniques predict what the outcome should
  be given a set of clinical characteristics
• These models are imperfect and being updated
  often (hopefully)

21
Improving Quality

• Continuous Quality Improvement method (CQI)
• To change the outcome, we have to change the
  process
• Focus is on the process and we use measurement to
  help understand the process and opportunities for
  changes
• Focus on reducing variations in the process

22
PDSA Method to Improve Quality

• Plan
• What are you going to do
• Do
• Just do it!
• Study
• Examine the process again
• Act
• Make necessary changes in mid-course
• And then start again!

23
Profiling and Report Cards

• Used to evaluate physician and hospital
  performance
• Examples
• VBAC and C-section rates
• CABG mortality and morbidity
• You will be profiled!

24
Leapfrog Group
25
Public Disclosure of Hospital Performance
Reference Hibbard et al. Health Affairs
200322(4)84
26
Pay for Performance

• The current payment system provides incentives to
  over use or under use care and to not improve
  care
• Does not reward efficiency!
• Medicare is rolling out systems to increase pay
  for MDs and health systems that provide better
  care

27
Class Project and Example
http//medinformatics.uthscsa.edu/ms4/
28

• There have been a number of complaints about the
  pharmacy turnaround time. Meds seem to take
  forever to arrive to the floor. The pharmacy
  does not seem to have any consistently large
  backlogs while they do have an occasional bad
  day if someone calls off. The Quality
  Improvement Team developed a simple flowchart to
  understand the process and they select several
  points for measurement. In July, 2003 the
  pharmacy responded by implementing some changes
  in medication verification and they feel as
  though they have reduced the time to verify a
  medication order. However, the complaints about
  long turnaround times have continued.

29

• Based on the data presented in the tab label
  DATA evaluate the processes associated with the
  medication turnaround time.
• 1. Have the pharmacy interventions made a
  difference in the pharmacys process?
• 2. What steps should be taken next to improve
  this process?

30
Medication Process
Physician writes paper order
Clerk enters order in computer
Clerk tubes order to pharmacy
Pharmacy verifies written order with computerized
order
Pharmacy dispenses medication
Medication given to patient
31

• Have the Pharmacy interventions made a difference
  in the Pharmacys process?
• Yes

32
(No Transcript)
33
2. What steps should be taken next to improve
this process?
34
(No Transcript)
35
How would you solve this problem?

• This facility added physician order entry and got
  rid of the steps involving the clerks

36
Conclusion

• There are enormous problems in the quality of
  health care that is delivered in the U.S.
• PDSA is one attempt to develop a method to
  improve that care
• Barriers to measuring quality of care include
• Defining appropriate outcomes
• Obtaining reliable data
• Developing valid risk-adjustment tools