Impact of antibiotic policies, guidelines and pathways of care on clinical, microbiological and econ

1
Impact of antibiotic policies, guidelines and
pathways of care on clinical, microbiological and
economic outcomes

• Gavin Barlow
• Consultant Physician
• Hull East Yorkshire Hospitals NHS Trust

2
Focus

• Hospital setting
• Positive studies, but there are negative ones!
• Robust study design

3
BackgroundWhat are we trying to achieve?
Knowledge/ attitudes
Intervention
Process
Outcomes
Target knowledge/ attitudinal deficits
External barriers
Process-outcome link
4
Can policies, guidelines and pathways improve
antibiotic prescribing?
What are we trying to achieve?
Appropriate antibiotic therapy
5
Can policies, guidelines and pathways improve
antibiotic prescribing?

• At least 7 controlled or ITS intervention studies
  have shown a change in antibiotic prescribing
• 2 policy, 4 guideline and 1 pathway based
• 2 of the 7 aimed to avoid under-treatment
• 3 of the 7 used other interventions (e.g.
  compulsory prescribing forms and stop orders)
• A variety of implementation strategies

6
Can policies, guidelines and pathways improve
antibiotic prescribing?
teicoplanin
pipericillin/tazobactam
ceftazidime
ciprofloxacin
Ceftazidime switched to pipericillin/tazobactam
Back to ceftazidime
Before
Intervention compulsory policy change Bradley et
al. J Antimicrob Chemother 1999 43261-266
7
Can policies, guidelines and pathways improve
antibiotic prescribing?

• Intervention Low guideline mailed usual QI
  methods Moderate
• educational session High reminders
  audit/feedback PDSA
• Yealy et al. Ann Intern Med. 2005143881-894

8
(No Transcript)
9
Intervention pneumonia management pathway
implemented using posters, educational sessions
and audit/feedback Barlow et al. 14th ECCMID,
Prague, May 2004
Before
After
Feedback
56
33
36
Intervention
32
Education
Winter 2002
Winter 2003
10
Can policies, guidelines and pathways improve
clinical outcomes?

• Intervention pneumonia guideline implemented by
  educational
• sessions, reminders, order sheets, and audit and
  feedback
• Dean et al. Am J Med 2001110451-57

11
Peterson et al. JAMA 20062951912-1920
12
Can policies, guidelines and pathways improve
microbiological outcomes?

• At least 3 convincing studies
• All policy based with 1 study also using
  individual case review and feedback
• Lack of data on non-target pathogens
• Lack of data on clinical outcomes

13
Can policies, guidelines and pathways improve
microbiological outcomes?
CDAD
Intervention
Resistant GNEB
Intervention consensus policy, pharmacy/ID
patient review and academic detailing 3rd
generation cephalosporins and aztreonam use
decreased from 24.7 to 6.2 DDD/1000
patient-days Carling et al. Infect Control Hosp
Epidemiol 200324699-706
14
Can policies, guidelines and pathways improve
microbiological outcomes?
National surveillance
MRSA
Intervention
National surveillance
Intervention
VRE
15
Can policies, guidelines and pathways improve
microbiological outcomes?
teicoplanin
pipericillin/tazobactam
ceftazidime
ciprofloxacin
Ceftazidime switched to pipericillin/tazobactam
Back to ceftazidime
Before
Intervention compulsory policy change Bradley et
al. J Antimicrob Chemother 1999 43261-266
16
Probability of remaining GRE free
Phase 2b
Phase 2a
Phase 3
Phase 1
n colonised Phase
1 43/75 57 Phase 2a 20/70 29 Phase
2b 5/59 8 Phase 3 21/58 36

Intervention compulsory policy change Bradley et
al. J Antimicrob Chemother 1999 43261-266
17
Intervention compulsory policy changedeMan et
al. Lancet 200018973-8
18
Intervention pneumonia management pathway
implemented using posters, educational sessions
and audit/feedback and emphasised both under- and
over-prescribingBarlow et al. Clin Microbiol
Infect 200612498-500
19
Did this change outcomes?
20
Can policies, guidelines and pathways improve
economic outcomes?
Cost of development
Cost of implementation
What are we trying to achieve?
Cost- effectiveness
Cost of evaluation
Change in provider outcomes
Change in patient outcomes
21
Can policies, guidelines and pathways improve
economic outcomes?
22
Intervention pneumonia management pathway
implemented using posters, educational sessions
and audit/feedbackBarlow et al. 10th FIS
Meeting, Cardiff, UK, November 2002
23
Intervention pneumonia management pathway
implemented using posters, educational sessions
and audit/feedbackBased on Barlow et al. 10th
FIS Meeting, Cardiff, UK, November 2003

• Cost per death prevented
• For no evaluation 3003
• For limited evaluation 16,632

Using data from Houck et al. Arch Intern Med
2004164637-44
24
Conclusions

• The good news
• Robust data to support the use of policies, etc.
  in improving process
• Good data to support the use of policies, etc. in
  improving microbiological outcomes, but need to
  link this to clinical outcomes
• The bad news
• Heterogeneous interventions and implementation
  and poor understanding of why interventions do or
  dont work
• Only one study has showed an improvement in
  clinical outcomes
• No high quality economic data

25
Conclusions

• The future
• Need for well designed intervention studies that
  measure process, all relevant outcomes and
  cost-effectiveness
• Need to better understand what works and why and
  in what context
• The danger
• Concern about patient outcome will continue to
  fuel over-prescribing
• Costly implementation of interventions that do
  not work