�

Dots, Data, and Decisions

Ontario Executive Patient Safety Series– Winter 2007

The Challenge• No needless deaths

• No needless pain or suffering

• No unwanted waiting

• No helplessness

• No waste

Source: http://www.ihi.org/ihi/Files/IMPACT/Spring2004LearningSession/2004-06-29/Berwick_Plenary.pdf

Much has Been Done …Trend in Age-Adjusted 30-Day In-Hospital Death Rate

Excludes NL, QC, BC

�

But Challenges Remain …

of Canadian adults report that they, or a family member, experienced a preventable “adverse event”

11.1%Overall

30-Day In-Hospital Death Rate Following New Heart Attack Admission

16.3%Highest

7.6%Lowest

Regional Differences 2002–2003 to 2004–2005, excluding QC & NL

Saving More Lives

Measurement for Improvement

�

Tracking Progress: Vital Signs

Team by Team

Projectby Project

BigDot

- Overall mortality trends

-Trends in care processes

- Intervention-level outcomes

-Tracking team’s care processes

What is HSMR? • Hospital Standardized Mortality Ratios

(HSMR) track changes in hospital mortality rates in order to:� Reduce avoidable deaths in hospitals� Improve quality of care

• Developed in the UK in mid-1990s by Sir Brian Jarman of Imperial College

• Used in hospitals worldwide (i.e. UK, Sweden, Holland and US)

Why Track Deaths?

• Significant numbers of deaths associated with adverse events in health care

• Definite, unique event

• Complements currently-available regional mortality rates for specific conditions � e.g. AMI & stroke survival rates

• Emerging evidence that institution-level death rates can motivate improvement and can be changed

�

Success stories

Tallahassee Memorial Hospital� HSMR in 2000 = 128.7� HSMR in 2004 = 89� Identified areas for improvement and

implemented targeted strategies in order to reduce HSMR to below average

Source: http://www.ihi.org

Reducing Deaths in WalsallHospital in the NHS

• In 2000, HSMR was 130� highest of all main acute hospitals in England

• Initially questioned data extensively

• Set up 7 clinical governance groups

• Implemented range of improvements

• Largest drop in mortality over past 5 years

http://bmj.bmjjournals.com/cgi/content/full/330/7487/329

Walsall change of HSMR

5060708090

100110120130140150

1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

HSM

R

First publication of HSMR Jan 2001

�

Saving More Lives in Walsall

5060708090

100110120130140150

1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

HSM

R

Represents ~ 290 fewer deaths

Understanding HSMR

What is the HSMR?

• HSMR is a measure designed to track changes in hospital mortality rates

• Based on diagnosis groups that account for 80% of deaths

• Adjusted for factors affecting mortality (e.g. mix of diagnoses, age, sex, length of stay)

Observed deaths

Expected deaths

HSMR= X 100

�

HSMR is easy to interpretEqual to 100� No difference between facility’s mortality rate

and average rate

• More than 100� Facility’s mortality rate is higher than the

average rate

• Less than 100� Facility’s mortality rate is lower than the

average rate

What Does Average Mean? (Results from Baker/Norton)

Extra hospital days associated with adverse events

Deaths among patients with preventable adverse events

Saving More Lives: Using HSMR

�

Progress to date

• Methodology developed and tested with Pioneer Group & experts

• ~ 350 sites from 125 organizations participating� First HSMR data sent in December 2005 � “Trend tracker” real-time quarterly HSMR updates

• Spreading the word, learning from early adopters

• Work in progress� Monthly (or more frequent) updates for larger facilities� Electronic HSMR� “Saved lives” calculation� Analysis of factors affecting results

How Are Hospitals Using HSMR?• Data validation

• Identifying opportunities� Clinical/improvement teams doing mortality reviews &

developing focused strategies (e.g. 2X2 matrix)� Data analysis (e.g. patients transferred in)

• Aligning with strategy� Board/senior team “big dot” indicator� Setting goals and tracking results

IHI: Raising the Bar ~ Setting Goals?

���� ������������������� ����������������������� ������ ��!��

��"����� ��������������� ��#����!����� �������������� �����!���$�������������!������%�����

��&��%��������� ��!������������������$���

��'�$�������������������������!�����������!��� ��������������(�����������������$������������ ��!��

��)�������� �����!������������������ ������������������������(�����

http://www.ihi.org/IHI/Topics/LeadingSystemImprovement/Leadership/EmergingContent/RaisingtheBarEmergingContent.htm

*

100,000 Lives Campaign

Crossing the Finish Line: IHI estimates 122,300 lives saved

Source: www.ihi.org

‘Big dots’ for other sectors An Example from Inpatient Rehabilitation

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8

Amputation of Limb

Neurological Conditions

Brain Dysfunction

Burns

Stroke

Spinal Cord Dysfunction

Arthritis

Pulmonary Disorders

Medically Complex

Debility

Major Multiple Trauma

Other Patient Groups

Cardiac

Pain Syndromes

Orthopaedic Conditions

Average Gain in Functional Status/Day

Source: National Rehabilitation Reporting System, 2004/05

IHI: Leadership Leverage Points

• Establish system-level aims for improvement

• Align measures, strategy, and projects

• Channel leadership attention

• Get the right team on the bus

• Make the CFO a quality champion

• Engage physicians� Engage organization in quality work of physicians

• Build improvement capability

+

The Road Ahead …