Improving the Effectiveness of NEWS,
Improving Care for Deteriorating Patients
John Welch, Consultant Nurse, Critical Care & Critical Care Outreach
Declaration of interest
http://www.nightingale-h2020.eu/
Subbe CP, Welch JR. Clin Risk. 2013. 19(1):6-11.
Record - Recognise - Report - Respond
The deteriorating patient journey
Potential problems
• Lack of knowledge of all inherent risk factors, frailty, etc
• Deciding, agreeing, documenting most appropriate care
• Incomplete vital signs, insufficiently frequent vital signs
• Inadequate understanding of abnormal vital signs
• Failure to effectively escalate abnormal vital signs
• Delayed response to escalation
• Delayed treatment
• Too much treatment – or too little
• Delayed re-checking that treatment has worked
Potential problems
• Lack of knowledge of all inherent risk factors, frailty, etc
• Deciding, agreeing, documenting most appropriate care
• Incomplete vital signs, insufficiently frequent vital signs
• Inadequate understanding of abnormal vital signs
• Failure to effectively escalate abnormal vital signs
• Delayed response to escalation
• Delayed treatment
• Too much treatment – or too little
• Delayed re-checking that treatment has worked
Subbe CP, Welch JR. Clin Risk. 2013. 19(1):6-11.
Record - Recognise - Report -
Respond
The deteriorating patient journey
Record & Recognise: The National Early Warning Score
NEWS scores & risk of arrest, ICU, death
Smith GB, et al. Resuscitation. 2013. 84(4):465-70.
In four years, ¾ of hospitals are using NEWS …
(36% are using some type of electronic system)
Record & Recognise: The National Early Warning Score
New news about NEWS vs Medical Emergency Team criteria
• Single centre study, 103,998 admissions • NEWS has higher specificity and generates less of a workload
Smith GB, et al. Crit Care Med. 2016. 44(12):2171-2181.
New news about NEWS and sepsis risk prediction
• Single centre study, 30,677 patients; 7,385 (24%) died or transferred to ICU
Churpek MM, et al. Am J Respir Crit Care Med. 2016 Sep 20.
New news about NEWS for non-elective medical & surgical patients
• Single centre study, 65,896 admissions • NEWS performed equally well, or better, for surgical as for medical patients
Kovacs C, et al. Br J Surg. 2016. 103(10):1385-93.
New news about NEWS for Emergency Department triage
• Single centre study, 500 patients; 27 (5.4%) with severe sepsis • The area under the curve for NEWS to identify risk of severe sepsis is 0.89
Keep JW, et al. Emerg Med J. 2016. 33(1):37-41.
Corfield AR, et al. Emerg Med J. 2014. 31(6):482-7.
• Multi centre study
• 2003 patients with sepsis
• ↑ NEWS = ↓outcomes (AUC 0.7)
Just one NEWS at ED triage is predictive
Initial NEWS & mortality in patients with sepsis
Initial NEWS 30-day mortality 0–4 5.5% 5–6 11.3% 7–8 13.3% 9–20 27.6%
NEWS: how to do it (my personal view)
1. “Tell and sell” the concept - tailored to the audience
2. Compare and contrast with the existing ‘track and trigger’ system
3. Set it out as a progressive development
4. Facilitate and support ward staff to input
5. Try out good ideas, e.g., with PDSA cycles
6. Challenge resistors: get data
7. Measure processes and outcomes
What’s new – on the front – “New Confusion” added to AVPU
(scores 3: needs urgent assessment)
18
What’s new – on the back – “New Confusion”:
think about delirium (pain, infection (sepsis), etc)
What’s new – on the front – “New Confusion” added to AVPU
(scores 3: needs urgent assessment)
20
What’s new – on the back – “New Confusion”:
think about delirium (pain, infection (sepsis), etc)
21
What’s new – on the front – modified approach to Sepsis
Subbe CP, Welch JR. Clin Risk. 2013. 19(1):6-11.
Record - Recognise - Report -
Respond
The deteriorating patient journey
89% of hospitals have Outreach
89% of hospitals have Outreach - 49% have 24/7 Outreach
89% of hospitals have Outreach - 49% have 24/7 Outreach
- 97% of teams are nurse only
Is it ok that UK teams are nurse only? “RRSs were associated with a reduction in hospital mortality and cardiopulmonary arrest.
Meta-regression did not identify the presence of a physician in the RRS to be significantly associated with a mortality reduction.” Crit Care. 2015. 19:254.
We’ve an efficient National Early Warning Score
Prytherch DR, et al. Resuscitation. 2010. 81(8):932-7.
Call 4 Concern© enables patients and families
to call for immediate help and advice when
they feel concerned that the health care team
has not recognised their own or their loved
one’s changing condition. The Outreach team
can be contacted directly if:
1. A noticeable change in the patient occurs
and the health care team is not recognising
your concern.
2. You feel there is confusion over what needs
to be done for the patient.
… and we’ve some ideas about more “soft” alerts
Odell M, et al. Br J Nurs. 2010;19(22):1390-5.
… and we’ve more ideas about more “soft” alerts
Odell M, et al. Br J Nurs. 2010 Dec 9-2011 Jan 13;19(22):1390-5.
1 Douw G, et al. Int J Nurs Stud. 2016;59:134-40.
More recent developments
Adjusted mortality rates at Queen Alexandra Hospital (top) and University Hospital Coventry (bottom). Schmidt PE, et al. BMJ Qual Saf 2015;24:10-20.
Measuring processes & outcomes
Four key questions
• Do you know how good you are?
• Do you know where you stand relative to the best?
• Do you know about variation in your system?
• Do you know how things change over time?
After Maureen Bisognano, IHI President/CEO.
‘Multi-disciplinary Audit EvaLuating Outcomes of Rapid Response’ = MAELOR
Outcomes Positive Negative
Transfer to ICU, or Theatre
1. Timely transfer, e.g., < 4 hours after the first trigger
2. Delayed transfer, e.g., > 4 hours after first trigger
Alive on ward 3. No longer triggering 4. Still triggering
Deceased 5. On terminal care pathway / with DNAR order
6. Following cardio-pulmonary arrest
Others 7. Alive with documented treatment limits / DNAR order
8a) Trigger from new pathology unrelated to previous call-out
8b) Chronic condition leading to continuous trigger
8c) Discharged from hospital
9. Outcome unknown
Morris A, et al. Crit Care Resusc. 2013;15(1):33-9.
Four key questions
• Do you know how good you are?
• Do you know where you stand relative to the best?
• Do you know about variation in your system?
• Do you know how things change over time?
After Maureen Bisognano, IHI President/CEO.
51 Hospitals Australia, Denmark, Netherlands, UK, USA
Resuscitation. 2016;107:7-12.
Ward referrals to UCH Outreach Team
• 34 referrals / 1000 admissions, 23% ICU • average NEWS at referral = 6.31 • 20.4% hospital mortality
Four key questions
• Do you know how good you are?
• Do you know where you stand relative to the best?
• Do you know about variation in your system?
• Do you know how things change over time?
After Maureen Bisognano, IHI President/CEO.
Completeness of NEWS scoring at UCH 91% of referrals had all seven vital signs and NEWS scores completed.
Accuracy of NEWS scoring at UCH 95% of referrals had accurate NEWS scores. There were two outliers.
Timeliness of referral to UCH Outreach Overall, 91% of referrals were timely, with only one outlier.
Use of SBAR
42
Timeliness of response from UCH Outreach 95% of patient referrals were responded to in a timely way. There are no outliers.
43
Timely transfer to ICU For patients transferred to ICU following referral to PERRT, 89% are transferred within 4 hours.
‘Multi-disciplinary Audit EvaLuating Outcomes of Rapid Response’ = MAELOR
Outcomes Positive Negative
Transfer to ICU, or Theatre
1. Timely transfer, e.g., < 4 hours after the first trigger
2. Delayed transfer, e.g., > 4 hours after first trigger
Alive on ward 3. No longer triggering 4. Still triggering
Deceased 5. On terminal care pathway / with DNAR order
6. Following cardio-pulmonary arrest
Others 7. Alive with documented treatment limits / DNAR order
8a) Trigger from new pathology unrelated to previous call-out
8b) Chronic condition leading to continuous trigger
8c) Discharged from hospital
9. Outcome unknown
Morris A, et al. Crit Care Resusc.2013;15(1):33-9.
Quality of Outreach Response 823 UCH Patient Emergency Response & Resuscitation Team referrals (2015)
Outcomes Positive PERRT +ve results
Negative PERRT –ve results
Totals
Transfer to ICU / Theatre
1. Timely transfer (<4 hours)
165 (89%)
2. Delayed (>4 hours)
21 (11%) 186
Alive on Ward 3. No longer triggering
494 (94%)
4. Still triggering
34 (6%) 528
Patient deceased 5. On terminal care pathway / DNAR order
6 6. Following CPR
7 13
Other Alive 7. with treatment limitations / DNAR
96 96
Totals 761 (92%)
62 (8%) 823
Positive and
negative outcomes
92%+ve 8%-ve
Of patients judged fit to stay on ward for active
treatment,494 of 528 were improved next day.
165 of 186 transfers to ICU occurred in <4
hours.
We can count and case-mix adjust deaths
Counting deaths … or, better, learning from them
Hogan H, et al. BMJ Qual Saf. 2012;21(9):737-45.
Hogan H, et al. BMJ. 2015;351:h3239.
Review sample of referrals to Outreach / unplanned transfers to ICU / arrests / deaths
49
Deteriorating Patients Care Bundle
Lawton R, et al. 369-80. BMJ Qual Saf. 2012; 21(5):369-80.
National, standardised, structured death reviews: framework of factors contributing to patient safety
Remember, nothing is certain, except …
Clark D, et al. Palliat Med.
2014;28(6):474-479.
(10,743 patients, 31/03/2010)
GSTT: Cardiac Arrests per month and Wards that have implemented AMBER
0
1
2
3
4
5
6
7
8
9
10
Sep-1
0
Oct-1
0
Nov-10
Dec-10
Jan-1
1
Feb-1
1
Mar
-11
Apr-11
May
-11
Jun-1
1
Jul-1
1
Aug-11
Sep-1
1
Month
Car
dia
c A
rre
st
0
2
4
6
8
10
12
14
16
18
20
Am
be
r W
ard
s
AMBER Wards: GSTT Arrest Data: GSTT
Critical Care Outreach
PDN meeting
Thanks to Adrian Hopper, GSTT
Outreach moves things along
GSTT: Cardiac Arrests per month and Wards that have implemented AMBER
0
1
2
3
4
5
6
7
8
9
10
Sep-1
0
Oct-1
0
Nov-10
Dec-10
Jan-1
1
Feb-1
1
Mar
-11
Apr-11
May
-11
Jun-1
1
Jul-1
1
Aug-11
Sep-1
1
Month
Car
dia
c A
rre
st
0
2
4
6
8
10
12
14
16
18
20
Am
be
r W
ard
s
AMBER Wards: GSTT Arrest Data: GSTT
Critical Care Outreach
PDN meeting
Thanks to Adrian Hopper, GSTT
Outreach moves things along
The following treatment plan should be used as clinical guidance and is not a substitute for ongoing consultation and shared decision-making wherever possible. The clinician should initial ONE of the patient’s priority boxes below, add relevant guidance in the large box and initial a CPR decision. The form must be signed, named and dated on the reverse.
Name:
Date of Birth: Hospital/NHS numbers:
Address:
1
This individual is FOR attempted CARDIOPULMONARY RESUSCITATION
Signature 6
This individual is NOT FOR attempted CARDIOPULMONARY RESUSCITATION Signature
If the patient dies in transit please take to: 6
Please provide clinical guidance on specific interventions that may or may not be wanted or clinically appropriate in community, hospital and critical care settings:
Provide details of other relevant care planning documents and/or documented wishes about organ/tissue donation (name and where held):
5
The priority is to get better. Please consider all treatment
to prolong life
Initials: .. . 4
The priority is to achieve a balance between getting better and ensuring good
quality of life. Please consider selected treatments
Initials: .. 4
The priority is comfort. Please consider all treatments aimed
at symptom control
Initials: .. 4
Turn over to complete this ECTP
Relevant information about the individual’s diagnosis, situation, ability to communicate, and reasons for the chosen plan.
3
Emergency Care & Treatment Plan
Date: __/__/____ 2
Designation - (Grade and specialty)
Print name & professional registration number
Signature Date and time
Senior Responsible Clinician
10
Plan review: If the individual’s condition changes (i.e. deterioration OR improvement) review the decisions on this ECTP. Document further conversations in box 8. If necessary, complete a new form, and write “CANCELLED” clearly across both sides of this form with signature and date. The decisions on this form should be reviewed specifically before any procedure during which abrupt deterioration or cardiac arrest may occur (e.g. endoscopy, cardiac pacing, angiography, surgery or anaesthesia). Make an agreed plan on whether or not to revoke temporarily the decisions on this form and, if so, on the treatments that will be considered if abrupt deterioration or cardiac arrest occurs. 11
Emergency contacts Name Telephone numbers Other relevant details
Welfare Attorney, Guardian etc.
Family/friend
GP
Lead Consultant
Specialist worker/key worker 12
Does the (adult) individual have capacity? (see guidance notes) YES NO
Do they have a valid advance directive or ADRT? YES NO
If so, record details in box 5
Do they have a representative with legal authority to make decisions? YES NO
(e.g. Welfare Attorney, Guardian, person with Lasting Power of Attorney for Health and Welfare)
If so, record their contact details and document details of discussion below. 7
The clinician signing this ECTP is confirming that these decisions: 1. have been discussed with and agreed with the individual; or 2. have been made in accordance with capacity law; or 3. in the case of a child, the person holding parental responsibility/court order. Date of discussion: __/__/____ Names of those present:
Full documentation of discussion can be found in:
Further conversations occurred on the following dates (state where details are recorded):
8
If there has been no shared decision-making with the individual, no shared decision-making with a
representative with legal authority to make decisions or no best-interests meeting for the individual who
lacks capacity, document a full explanation and a clear plan to address this in the clinical records.
Summarise the reason (e.g. describe any potential to cause harm) here:
9
and we’ve a new, national combined ECTP/DNACPR form
What’s new in NEWS 2
• Re-ordering of the chart layout
• Recording of oxygen therapy
• Consideration of chronic respiratory disease
• Highlighting of new confusion / delirium
• Sepsis …
What else?
Thinking about the whole system
Clear policy & procedure; new NEWS chart(s)
Monthly e-audits → Care thermometer
wards ‘know how they’re doing’
Frontline comms: ‘Message of the week’
PERRT training
Ward safety huddles each shift
58
Continuous improvement …
Key: ward staff – and the patients – know
how they’re doing more or less in real time
Huddles to improve Teamwork and
Communication
.
Learning from serious incidents
62
AAR: The Four Steps What was EXPECTED What ACTUALLY happened
WHY there was a difference What can be LEARNED
Before the event, what was the
objective, plan or expected outcome?
It could be a shared plan, a formal
agreement, a guideline, a personal
expectation, or simply regular practice.
After event, each participant describes:
What they did, saw or experienced
During the event.
Explore the facts, while acknowledging the
perspective and feelings of others.
Why was there a difference between
the expected outcome and the reality
of the moment?
Check if expectations were properly
shared, and what constraints on people,
time or resource prevented expectation
being realised.
Learning is the prime action within an
AAR. What will be different next time?
It may be a change in practice or policy, or
a change in attitude, behaviour, shared
understanding or greater insight. Direct the
collective wisdom to improving future
performance.
Thinking about BARRIERS
Knowledge: Confusion over how to treat complicated patients
(fluid balance, long term in-patients)
Social Influences: Lack of communication: ‘Is this patient on the pathway or not?’, conflict between Drs
and Nurses
Beliefs about consequences: Fear of harming patients with Sepsis Six,
lack of confidence in the evidence
Steinmo S, et al. Implement Sci. 2015;10:111.
and LEVERS
Memory and Attention: Sepsis Six ‘branding and marketing’,
plus prompts and reminders
Environment: Materials and resources immediately available
Social influences: Superiors’ commitment; reciprocal feedback ‘It’s our pathway and
we’re being listened to’.
Beliefs about consequences: seeing health improve immediately,
following-up specific patients
Steinmo S, et al. Implement Sci. 2015;10:111.
Focused training
Technological aids in crises
and data and analytics
is something else
- derived from analysis of 5 million patient encounters in a wide range of hospitalsepsis is likely with
- infection and ≥2 of RR ≥22, SBP ≤100, altered mentation
-
Singer M, et al. JAMA. 2016;315(8):801-10.
Artificial Intelligence 4, Human Champion 1
Subbe CP, Welch JR. Clin Risk. 2013. 19(1):6-11.
Record - Recognise - Report - Respond
The whole system needs to be right …
Four key questions
• Do you know how good you are?
• Do you know where you stand relative to the best?
• Do you know about variation in your system?
• Do you know how things change over time?
After Maureen Bisognano, IHI President/CEO.
Thanks!