Clinical Audit Theory and Practice

1
Clinical Audit Theory and Practice

• A/Prof Graeme Hart
• Clinical Director
• Austin Centre for Applied Clinical Informatics

2
What is Quality in Health?

• ..degree to which health services for individuals
  and populations increase the likelihood of
  desired health outcomes and are consistent with
  current professional knowledge
• Boyce et al 1997
• and reduces the probability of undesired
  outcomes
• US Office of Technology Assessment

3
Service Delivery Problems

• Over Use Providing Service when Risk exceeds
  Benefit
• Under Use Failure to Provide Proven Care
• Mis Use Avoidable Complications of Appropriate
  Care

Chassin 1998
4
Dimensions of Quality

• EFFECTIVENESS
• What is the right thing to do
• Based on appropriate research
• APPROPRIATENESS
• Was the right thing done?
• PERFORMANCE
• Was the right thing done properly and well?
• ( safety and technical quality)
• Was the OUTCOME satisfactory?
• (including acceptability)

Fletcher M, The Quality of Australian Healthcare
Current Issues and Future Directions. Commonwealth
Dept Health and Aged Care
5
Whose audit ?
Funders / Government / Insurers
Patient
External Clinical Peer Group
Doctor
Clinical Department
6
Types of data collection and analysis

• Formative
• presentation of data influences process and
  practice
• Summative
• presentation provides an overview of aspect of
  practice, usually for external party
• Archival
• Collated and present for the record
• Others
• scientific, exploratory, legal system, financial

7

• Formative
• Data should be collected and analysed to allow us
  to shape the systems and processes.
• Not merely collected to reassure us or document
  failure.

8

• Charts can provide a signal
• Its about understanding the process, choosing and
  grading KPI based on explicit criteria
• Having a process to respond to signals if they
  occur

9
MET call rate
10
Control Chart ICU adjusted mortality rate
95 confidence intervals
Predicted mortality
Observed mortality
11
(No Transcript)
12
Clinical Audit

• Significant amount of administrative data held
  but information not widely available
• Age, Sex, Dates
• Diagnoses, Procedures, Complications (ICD10-AM
  coded by HIS)
• CMBS procedures, complications
• Proceduralists Unit, Consultant, Registrar
• Add ICU stay and ventilator hours, tracheostomy
• Hospital Outcome.

13
Query example

• Admission Date Range
• Age range
• Sex
• Admission Diagnosis
• Surgeon
• Procedure
• Complication
• Length of Stay
• Unplanned ICU admission
• Outcome Status

14
Query example

• Between Jan 2000 and June 2003
• gt 65 yrs
• All
• Cholecystitis
• Jones
• Laparoscopic Cholecystectomy
• Bile Leak
• All

15
(No Transcript)
16
(No Transcript)
17
(No Transcript)
18
(No Transcript)
19
CIs of SMR sites 21-40

21n
21bc
21bca
21tyl
21ltyl

22n
22bc
Sites
1
SMR
0
1.63

20
95 CIs of ICU SMR ranks

50
63
53
48
17
56
95
Sites
25
49
73
1
98
Ranks

21
Baseline incidence of cardiac arrest
(Sources Crit Care Med 1994, J Am Coll Cardiol
1994, MJA 2000)
Arrests / 10,000 admissions
22
Baseline in-hospital mortality
Sources MJA 2000, Am J Med Qual 2000
Deaths/10,000 admissions
23
Surgical performance
Source N Engl J Med 20023461128-37, Austin
Medical Records Dept.
Patients gt 65 years of age
Mortality
24
Surgical performance
Source N Engl J Med 2002, CMAJ 2003, Austin
Medical Records Dept.
For patients gt 65 years of age Average of last 5
years
Mortality
25
(No Transcript)
26
(No Transcript)
27
(No Transcript)
28
(No Transcript)
29
Austin Health requires a viable and sustainable
clinical audit system

• Current audit
• large unquantified proportion of many clinicians
  time.
• Local needs are fiercely advocated
• local solutions strongly protected
• Institutional blind spot
• Not proactive nor unified
• Potential benefits from co-ordinated approach and
  single platform
• Needs structure, supervision, software,
  governance.
• Goal - system meeting local needs but provide
  to many and overview for the Hospital at large.

30
The tools we acquire should have

• Security enabling
• Ability to progressively link across multiple
  feeder databases clinical TRAK/PAS/labs/departmen
  tal databases of varying architectures
• Simple graphical user interface to ensure
  maximum uptake and minimum training requirements
• pre defined generic reports with various data
  cubes
• extract KPIs on a regular basis and set up time
  series analysis for improvement tracking.
• Statistical capability
• Extracts for training logs of registrars
• Web enablement

31

• Audit - include all aspects of clinical care,
• clinician performance,
• procedural outcomes,
• assessment of new clinical programs such as HARP,
  HITH.
•  Clinical Governance processes are severely
  restricted without such tools.
• Audit should map to financial outcomes and other
  KPIs available through non clinical systems such
  as financial, PAS.
• Statistical Epidemiological Support