dr gareth kantor 1

Best Care…Always! Campaign

Global Forum on Bacterial Infec3on New Delhi ; October 4, 2011

Dr Gary Kantor

A loose coalition of stakeholders can initiate / sustain system strengthening for patient safety. The process began with antibiotic stewardship, and was driven by a private funder, hospitals and professionals. We are exploring new ways of building will, generating and sharing ideas and filling the execution gap ….changing the system.

Global Epidemic of Harm in Hospitals

3

Adverse events in 9 – 18% of admissions

~ 50% preventable

2.5 – 7.5% are fatal

Qual Safety in Health Care 2008;17:216-223

NEJM Nov 25, 2010

n=795 3 hospitals

Preventable Harm: 1 in 3 Hospital Patients

4

#1. procedures

Health Affairs, 30, no.4 (2011):581-589

Voluntary reports are 1% of events

#3. infection

#2. medications

Hospital-Acquired Infection •  World

–  1.4 million patients affected / day

•  Developed countries –  Hospital incidence up to 10% –  USA: 100,000 deaths

•  Developing countries –  3 x higher –  S Africa 9.7% prevalence; 28.6% ICU

A Duse. SA-HISC study (unpublished) Allegranzi B; Lancet 2010:61458

5

JAMA 2009;301(12):1285-1287 Lancet 2008;372(9651):1719-1720

Pronovost P. NEJM Dec 2006

6

Median rate of infection: ZERO!!

CLABSI rates ↓ by 66% Better than 90% of US ICUs 1,500 lives and $$$ saved Sustained > 3 years

96 ICUs

7

maximal barrier precautions chlorhexidine skin antisepsis optimal catheter site selection daily review of line necessity

EVIDENCE

+ Daily goals sheet

+ Unit-based safety program

ü ü ü ü

CLABSI Bundle

Pronovost P. NEJM Dec 2006

Checklists

8

+ Antibiotic Stewardship

CAUTI CLABSI

SSI VAP

9

Expert Panel

Public Sector

Professionals

Private Hospitals

Sponsors

Improvement Scientists

BUILDING WILL

evidence-based medicine Evidence-based IMPLEMENTATION

Effective changes that lead to an improvement

10

Subject Matter Knowledge

Improvement Science

“what”

“how”

“where”

from

to

11

at least 1 intervention

202 Hospitals enrolled

“The Method for Improvement” = the Scien3fic Method

12

Not protocols

Not “recipes”

Decrease CLABSI to 0 or 300 patient days between by June 2010

AIMS 10 DRIVERS 20 DRIVERS

Reliable implementation of the insertion bundle

Reliable implementation of the maintenance bundle

Reliable identification of a CLABSI

Improve safety culture through multidisciplinary team working and communication

Bundles known by staff

Insertion bundle (HII)

Reliable data collection process – infections & device days

Line insertion documented in notes

Reliable data collection

Reliable Definition – CLABSI

Clinical leadership rounding to view lines

Maintenance bundle (HII)

HAI surveillance data and bundle compliance shared with staff

Reliable data collection

95% compliance with insertion bundle and each element of bundle 95% compliance with maintenance bundle and each element of bundle

30% reduction in antibiotic overuse

Optimal antibiotic use in 80% of patients

AIMS 10 DRIVERS CHANGE CONCEPTS

CAUTI bundle

SSI bundle

CLABSI bundle

VAP bundle

Antibiotic form

Clinical pharmacist review

Path lab hotline

Periodic review for cessation, route, reason for treatment

Prescriber access to knowledge and data

Cost reports

Prompt initiation, for defined reasons

Stable / decreased antibiotic resistance

Resistance reports

Prevention of hospital-acquired infection

*Prevent SSI, CLABSI, VAP and CAUTI

INTERVENTIONS

Day 3 and Day 7 review

Separate AB prescribing from other Rx

Info on how to Rx

Info on what it costs

↑ availability of first dose

Antibiotic ward stock

AB Bundles

*Interventions already associated with the BCA campaign

OUTCOME MEASURE

10 PROCESS MEASURE 20 PROCESS MEASURE

% with compliance to all bundles (“optimal use”)

% receiving timely antibiotics for prevention or treatment – first antibiotic prescribed during hospital course

% compliance with each Inception bundle element:

1.  <2 hrs from order → admin (treatment)

2.  Prophylaxis within 1 hr of incision

% overall compliance with Day 3 Bundle for the first antibiotic prescribed during hospital course

% compliance with each Maintenance bundle element:

1.  Treatment not prophylaxis 2.  State antibiotic indication or stop 3.  Culture(s) ordered or done 4.  Reassess drug choice

% overall compliance with Day 7 Bundle for the first antibiotic prescribed during hospital course

% compliance with each Maintenance bundle element:

1.  Stopped or re-ordered 2.  Conversion from IV to oral or N/A

Expert and Planning

Group formed

Learning session

1

Learning session

2

Repeated improvement

cycles:

Repeated improvement

cycles:

Learning session

3

18 -24 months

Mentoring and support

16

IDEAS

5.57

3.85

5.66

4.21

7.17

4.95 5.22

2.01 2.10 2.01

3.22 2.93

3.36

2.58

3.12

2.46

3.48

2.90 2.98

2.08 2.33

2.17

1.57

2.05

1.67

0.57 0.85

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

CLABSI -‐ Infec9on Rates Per 1000 Central Line Days Life Healthcare Group -‐ Oct 2008 to Sept 2010

CLABSI-‐Rate Mean

Upper control limit Lower control limit

Repor3ng system and training on BCA in all acute hospitals Cross func3onal workshops in

ICU's to implement bundle compliance ac3ons and increase involvement of Unit Managers

17

n=41 hospitals

HAI : CLABSI RATE

4.95 5.22

2.01 2.10 2.01

3.22

2.93

3.36

2.58

3.12

2.46

3.48

2.90 2.98

2.08 2.33

2.17

1.57

70%

87% 83% 83%

78%

83% 85%

83% 85% 84% 84%

88% 87% 87% 87% 87% 90%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

-‐

1.00

2.00

3.00

4.00

5.00

6.00

Mar-‐09

Apr-‐09

May-‐09

Jun-‐09

Jul-‐0

9

Aug-‐09

Sep-‐09

Oct-‐09

Nov-‐09

Dec-‐09

Jan-‐10

Feb-‐10

Mar-‐10

Apr-‐10

May-‐10

Jun-‐10

Jul-‐1

0

Aug-‐10

Central Line Associated Blood Stream Infec9ons -‐ Bundle Compliance and Infec9on Rate Mar 09 -‐ Aug 10

Infec3on Rate

CLABSI

Central Line Associated Blood Stream Infections - CLABSI

18 n=41 hospitals

BCA : COMPLIANCE : SSI

Bundle Compliance to SSI Period: JUNE-10

57%

89% 89%

96%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

If hair is removed, it is only done withclippers or dipilatory cream

Antibiotics are given within an hour ofincision

Glucose is maintained above 4 andbelow 8 after the initial post operative

assessment in ICU

The patients temperature ismaintained at >36.5 and <37.2 afterthe initial post operative assessment

in ICU

19

313029282726252423222120191817 •161514 •1312 • •11 • •10 •9 • •8 •7 •6 • • • • •5 •4 • • • •3 •2 • •1 • • •0 • •

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 1 2 3 4 5 6 7 8 9 10 11 12

Neurosurgery ICU -‐ Steve Biko Academic Hospital -‐ PretoriaAugust -‐

Infection

9 10 11 12

11 17 4 414 2 1 12 9 4

5

Central Line Associated Bloodstream Infections

Month: April -‐ July 2010

2-‐Oc

t

22-‐Aug

26-‐Aug

7-‐Sep

24-‐Sep

28-‐Sep

9-‐Jul

15-‐Ju

l

29-‐Ju

l

31-‐Ju

l

1-‐Au

g

13-‐Aug

1-‐Jun

10-‐Ju

n

18-‐Ju

n

24-‐Ju

n

1-‐Jul

3-‐Jul

2-‐May

7-‐May

8-‐May

11-‐M

ay

22-‐M

ay

26-‐M

ayDate of infections

4-‐Ap

r

14-‐Apr

15-‐Apr

26-‐Apr

26-‐Apr

Days between infections 10 1 11 0 6 5

212 13

8 6 7 2 6 61 3 11 4 6 9

Days between infections

Data Element 1 2 3 4 5 618 19 20 21 22 16 7 8 9 10 11 7 83 4

Started with

CLABSI Bun

dle

14 15 16 17

Possible contam

inant

Infection 3 days after adm

ission

20

Days Between Infection

No incident

New incident More than 1

incident Poor data

21

Visual Measurement

At the same time every day the Unit manager counts devices in use in the ward

Solving the Denominator Problem

23

www.bestcare.org.za

[email protected]

Not just infections

Not just bundles

24

Everyone in healthcare has 2 jobs

1. Doing the work 2. Improving the work!

All improvement requires change (though not all change is an improvement) Changing:

How and why we measure Methods (of improvement) Our sense of responsibility Leadership Organisations and culture The Health industry

25

How and why we measure

OLD

International data

Counts

Bar graphs

Data for head office / ministry

Individual measures

NEW

Our data

Rates

Run charts (over time)

Measurement for frontline staff

Measures across systems

26

Modified from: D van den Bergh, Netcare Hospital Group

27

BMJ Qual Saf 2011;20:46e51

Rules for iden3fying non-‐random signals

“Run Chart”

27

How we improve

OLD

Audit and inspection (QA)

Checklists for checking

Writing more protocols

“Spray and pray”

NEW

PDSA cycles

Checklists as aids

Focused interventions

Improving critical elements one a time

28


Taking Responsibility OLD

“it doesn’t happen here”

“we already do that”

Can’t do

Accept the inevitable

Victim of limitations

NEW

Knowing the facts

Acknowledging we may not

“if they can so can we”

Persistence

Building skills

29


Clinicians

Skeptical and critical

“this might work”

“worth trying”

“how can we support you”

“I would like to initiate”

30

Leadership OLD

It’s up to the doctors

It’s up to the nurses

It’s up to the Infection

Prevention Practitioners

It’s up to the Infection Control Committee

NEW

Active involvement of senior leadership

“Exco”

31


The Culture OLD

Blaming and punishing

Who (people)

Helping

NEW

Learning and curious

Why (system)

Capacitating (mentors)

32


Health Sector NEW

Collaboration

Sharing

Interconnected systems

Public learns from Private

Private learns from Public

Public-private partnership

OLD

Competition

Secrets

Private vs public


33

Antibiotic Stewardship OLD

Passive observers

No interventions

No measures

Defensiveness

Pilots

NEW

Actively seeking solutions

Identifying opportunities

First level utilisation data

Working together to deal with it

Multiple sites testing change

34


35

www.bestcare.org.za

The Changing View of Quality

We are perfect!

NO ACTION

Get rid of the bad apples

M&M

Quality Assurance

REACTION

Incident reporting

“Standards”

36

The Changing View of Quality

We are perfect!

INACTION

Get rid of the bad apples

System thinking

M&M

Quality Improvement

“Quality Assurance”

REACTION PRO-ACTION

“Quality” Safe

Effective Timely

Equitable Patient-centred

Efficient

Improvement Science

Incident reporting

“Standards”

37

Dashboard of Quality Measures

Generalisable scien9fic evidence

Improvement Science and Knowledge Systems Combine to Produce Improvement

High performance (measured)

Plans for change

Par9cular context

Execu9on

+

+

+

Qual Saf Health Care 2007;16:2–3

Patients get “recommended care” ~ 50% of the time

“control context”

“include time”

“local processes, habits, traditions”

standardisation, forcing functions, education, etc

“drivers of change”

38

dr gareth kantor 1

Documents