Measuring harm in healthcare

Report
Measuring harm in
healthcare
Our demographics are changing…
Our tolerance is changing….
The burden is growing…….
Measurement is complex
Hospital Episode Statistics
Mortality (in hospital)
Decubitus ulcer
Length of in-patient stay
VTE
30 day Readmission
http://www.nrls.npsa.nhs.uk/
http://www.nrls.npsa.nhs.uk/resources/?EntryId45=135153
Are we improving?
The 6 million dollar question
MRSA Infection Rates
MRSA All England Rate per 1,000 Bed Days
u' chart
Set 5: UCL = 0.02, CTL = 0.01, LCL = 0.00 (51 - 70) (Lloyd Nelson option)
Inspected Mean = 3,041.33, Counts Mean = 33.94
0.10
0.09
0.07 to 0.01 per 1000 bed days = 86% 
UCL
0.08
MRSA Rate per 1,000 Bed Days
0.07
CTL
0.06
LCL
0.05
UCL
0.04
CTL
0.03
LCL
UCL
CTL
UCL
0.02
LCL
0.01
CTL
UCL
LCL
CTL
LCL
01
/0
4/
2
01 00
7
/0
7/
2
01 00
7
/1
0/
20
01
07
/0
1/
2
01 00
8
/0
4/
20
01
08
/0
7/
2
01 00
8
/1
0/
2
01 00
8
/0
1/
20
01
09
/0
4/
2
01 00
9
/0
7/
20
01
09
/1
0/
2
01 00
9
/0
1/
2
01 010
/0
4/
2
01 01
0
/0
7/
2
01 01
0
/1
0/
20
01
10
/0
1/
2
01 01
1
/0
4/
20
01
11
/0
7/
2
01 01
1
/1
0/
20
01
11
/0
1/
2
01 01
2
/0
4/
2
01 01
2
/0
7/
20
01
12
/1
0/
2
01 01
2
/0
1/
20
13
0.00
Month
Data sourced from the Health Protection Agency (mandatory surveillance of MRSA, Trust apportioned cases based on 167 NHS Acute Trusts
in England) and HES online (calculated hospital bed days), calculated for the period April 2007 – October 2012. The data is presented as a rate
(U Prime Chart) of MRSA bacteraemia per 1,000 hospital bed days (based on HES number of episodes). The mean rate for the latest re-based
period is 0.01 per 1,000 FCE’s, an 86% decrease from the first period centre line. During the reported period there has been a decrease
(before and after, first (Aug-Oct) and last three data points) of 2,558,522 bed days.
C.Diff Infection Rates
C.Diff All England Rate per 1,000 Bed Days
u' chart
Set 5: UCL = 0.20, CTL = 0.17, LCL = 0.15 (56 - 70) (Lloyd Nelson option)
Inspected Mean = 3,006.32, Counts Mean = 521.75
1.2
1.1
0.79 to 0.17 per 1000 bed days = 78% 
1.0
UCL
C.Diff Rate per 1,000 Bed Days
0.9
0.8 CTL
0.7
LCL
0.6
UCL
0.5
CTL
0.4
LCL
0.3
UCL
CTL
UCL
LCL
CTL
0.2
UCL
LCL
CTL
LCL
0.1
01
/0
4/
20
07
01
/0
7/
2
01 007
/1
0/
20
07
01
/0
1/
20
08
01
/0
4/
2
01 008
/0
7/
20
08
01
/1
0/
20
08
01
/0
1/
2
01 009
/0
4/
20
09
01
/0
7/
20
09
01
/1
0/
2
01 009
/0
1/
20
10
01
/0
4/
2
01 010
/0
7/
20
10
01
/1
0/
20
10
01
/0
1/
2
01 011
/0
4/
20
11
01
/0
7/
20
11
01
/1
0/
2
01 011
/0
1/
20
12
01
/0
4/
20
12
01
/0
7/
2
01 012
/1
0/
20
12
01
/0
1/
20
13
0.0
Month
Data sourced from the Health Protection Agency (mandatory surveillance of Clostridium Difficile, Trust apportioned cases based on 167 NHS
Acute Trusts in England) and HES online (calculated hospital bed days), calculated for the period April 2007 – October 2012. The data is
presented as a rate (U Prime Chart) of C.Diff per 1,000 hospital bed days (based on HES number of episodes). The mean rate for the latest rebased period is 0.17 per 1,000 FCE’s, a 75% decrease from the first period centre line. During the reported period there has been a decrease
(before and after, first (Aug-Oct) and last three data points) of 2,558,522 bed days.
The proportion of patients who die in hospital
19%
Proportion of patients who die (%)
Seasonal Adjustment
1.88%
1.52%
Months
Data sourced from Hospital Episode Statistics (HES) for the period April 2007 – October 2012. The data is presented in proportion (p prime
chart) of patients in England that die within hospital (both emergency and elective), excluding non-elective/non-emergency e.g. maternity
and blank codes. The denominator is the number of hospital admissions. The mean crude rate for the latest rebased period is 1.51%,
although this does not take into account seasonal peaks. This data does not highlight the wide variation in mortality rates between
organisations. During the reported period monthly admissions have remained stable at approximately 1 million per month. However between
2007 (Aug-Oct) and 2012 (Aug-Oct) there has been an increase in admission of 15% (a difference of 427,691).
The proportion of patients who die in hospital – NHS QUEST network
Data sourced from Hospital Episode Statistics (HES) for the period April 2007 – October 2012.
The data is presented in proportion (p prime chart) of patients in England that die within
hospital (only emergency admissions), excluding non-elective/non-emergency e.g. maternity
and blank codes. The denominator is the number of hospital admissions.
QIPP Safe Care
What about these harms?
Submissions of data in the NHS Safety Thermometer
200000
700
180000
600
500
140000
120000
400
100000
300
80000
60000
200
40000
100
20000
0
0
Q1
Q3
Pre-work
Q4
Q5
Safety Express
Q6
Q7
Policy
Q8
CQUIN
Q9
Q10
Q11
CQUIN implemented
Organisations
Patients
160000
Scale up to a national data collection
How many patients have a pressure ulcer?
How many people have a new
pressure ulcer?
Is there a difference between settings?
What category is causing the highest
burden?
How much variation is there?
Hospital Group New Pressure Ulcers
Number of patient affected by a New Pressure
(Cat 2-4)
Number of patient affected by
New Pressure Ulcer
(Cat 3-4 only)
Percentage of patients affected by a
New Pressure Ulcer
(Cat 2-4)
Percentage of patients affected by a
New Pressure Ulcer
(Cat 3-4 only)
Organisation
Incident
Reporting
Safety
Thermometer**
Incident
Reporting
Safety
Thermometer
Incident
Reporting
Safety
Thermometer
Incident
Reporting
Safety
Thermometer
The Newcastle Upon Tyne
Hospitals NHS Foundation Trust
800
1844
---
491
0.6%
1.4%
---
0.4%
Sheffield Teaching Hospitals
NHS Foundation Trust
243
2098
---
270
0.2%
1.6%
---
0.2%
Central Manchester University
Hospitals NHS Foundation Trust
842
1215
---
273
0.6%
0.9%
---
0.2%
Oxford University Hospitals
NHS Trust
248
1948
---
209
0.2%
1.6%
---
0.2%
Imperial College Healthcare
NHS Trust
22*
1041
---
38
0.0%
0.8%
---
0.0%
Cambridge University Hospitals
NHS Foundation Trust
153
629
---
55
0.2%
0.6%
---
0.1%
University College London
Hospitals NHS Foundation Trust
222
428
---
142
0.3%
0.5%
---
0.2%
Guy's and St Thomas' NHS
Foundation Trust ***
80
613
---
68
0.1%
0.6%
---
0.1%
King's College Hospital NHS
Foundation Trust
147
1328
---
189
0.2%
1.3%
---
0.2%
501*-
266
---
---
0.7%
0.3%
---
---
University Hospitals
Birmingham NHS Foundation
Trust
*
***
***
---
Cat 2 Pressure Ulcers not included
Cat 1 Pressure Ulcers included
Extrapolated figures based on full years HES (2011/12) admissions (inpatient spells) for Shelford group organisations and NHS ST national data for financial year 2012/13
Community reporting of pressure ulcers is proportionally higher than Acute reporting (Acute 80 & Community 717)
No data available
What does all of this really
mean?
186, 617
People harmed from pressure ulcers
developing in Hospital in England alone
149, 293
Could be avoided
No national reduction in 2012…
Commit to Act Now
Tweet ideas #notonmywatch
What are we learning from the evaluation?
Issues common to data
collection systems:
• Connection to purpose
• Data collection systems
• Use of data
http://harmfreecare.org
Thank you

similar documents