Care bundles involve grouping several evidence-based interventions together that are aimed at improving patient outcomes when implemented collectively rather than individually. The ventilator bundle includes interventions like DVT prophylaxis and head of bed elevation. Compliance with care bundles is measured using both process measures like completion rates of individual elements as well as outcome measures. Studies show care bundles can improve outcomes for conditions like sepsis when bundle elements are reliably completed for all applicable patients. Ongoing auditing of compliance and outcomes is needed to demonstrate the sustained impact of care bundles.

1. UUssee ooff CCaarree BBuunnddlleess iinn
CCrriittiiccaall CCaarree
PPaauull FFuullbbrrooookk,, AAuussttrraalliiaa
PPrrooffeessssoorr ooff NNuurrssiinngg,, AAuussttrraalliiaann CCaatthhoolliicc UUnniivveerrssiittyy
NNuurrssiinngg DDiirreeccttoorr RReesseeaarrcchh,, TThhee PPrriinnccee CChhaarrlleess HHoossppiittaall,, BBrriissbbaannee
AAddjjuunncctt PPrrooffeessssoorr,, UUnniivveerrssiittyy ooff TTaassmmaanniiaa
NNaattiioonnaall VViiccee PPrreessiiddeenntt AACCCCCCNN

2. Outline
Background and definition
The ventilator bundle
Measures
– Compliance and reliability
The sepsis bundles
– Compliance
– Outcomes
– Current trials

3. Background: EBP
• It is imperative that those working in critical care
environments examine their practice to ensure that
it is evidence-based and of a high quality.
• The importance of knowledge-based practice is
constantly emphasised, with the expectation that
evidence-based decision-making and practice is at
the heart of all healthcare (Bonell, 1999).
• Evidence-based practice emphasises the use of
existing research.

4. CCaarree BBuunnddllee:: DDeeffiinniittiioonn
AA ssmmaallll sseett [[nnoo mmoorree tthhaann 55]] ooff
eevviiddeennccee--bbaasseedd iinntteerrvveennttiioonnss ffoorr aa
ddeeffiinneedd ppaattiieenntt sseeggmmeenntt//ppooppuullaattiioonn aanndd
ccaarree sseettttiinngg tthhaatt,, wwhheenn iimmpplleemmeenntteedd
ttooggeetthheerr,, wwiillll rreessuulltt iinn ssiiggnniiffiiccaannttllyy bbeetttteerr
oouuttccoommeess tthhaann wwhheenn iimmpplleemmeenntteedd
iinnddiivviidduuaallllyy
RReessaarr eett aall.. IIHHII IInnnnoovvaattiioonn SSeerriieess 22001122..

5. CCaarree bbuunnddlleess
· The ‘bundle’ concept was developed
originally in 2001 by IHI
· First described in nursing literature (Fulbrook
Mooney, 2003)
· Goal: to improve critical care processes
· Each element relatively independent
· Used within a defined population in one
location

6. Evidence-based practice
Care bundles provide a method for establishing best
clinical practice, which is evidence-based.
Individual components of each care bundle should be
well defined and based on strong science; usually
level one or two research.
By grouping evidence-based practices together,
within a single clinical protocol that guides patient
management (or process), the overall quality of care
given to critically ill patients will improve.

7. Levels of evidence
Level of evidence Criteria
1 Strong evidence from at least one systematic review of
multiple well designed randomised controlled trials
2 Strong evidence from at least one properly designed
randomised controlled trial of appropriate size
3 Evidence from well designed trials without randomisation,
single group pre-post, cohort, time series or matched case
control studies
4 Evidence from well designed experimental studies from
more than one centre or research group
5 Opinions of respected authorities, based on clinical
evidence, descriptive studies or reports of expert
committees
(after Moore et al., 1995)

8. There should
be sufficient
evidence to
support each
element of the
bundle
Each element
should be
applied to
most, if not all,
patients

9. TThhee vveennttiillaattoorr ccaarree bbuunnddllee
· DDeeeepp vveeiinn tthhrroommbboossiiss ((DDVVTT)) pprroopphhyyllaaxxiiss
· GGaassttrriicc uullcceerraattiioonn pprroopphhyyllaaxxiiss
· HHeeaadd ooff bbeedd eelleevvaattiioonn ((3300oo))
· SSeeddaattiioonn hhoolldd ((sseeddaattiioonn vvaaccaattiioonn)) aanndd aasssseessssmmeenntt ooff
rreeaaddiinneessss ttoo wweeaann
IInn 22001100,, aa 55tthh eelleemmeenntt wwaass aaddddeedd::
– DDaaiillyy oorraall ccaarree wwiitthh cchhlloorrhheexxiiddiinnee
GGoooodd eevviiddeennccee ooff aassssoocciiaattiioonn bbeettwweeeenn VVCCBB
iimmpplleemmeennttaattiioonn aanndd VVAAPP rreedduuccttiioonn ((LLaawwrreennccee FFuullbbrrooookk,, 22001111))
FFeeeeddbbaacckk iimmpprroovveess ccoommpplliiaannccee bbuutt HHOOBBEE ccoommpplliiaannccee
rreellaattiivveellyy ppoooorr ((LLaawwrreennccee FFuullbbrrooookk,, 22001122))

10. PPrroocceessss aanndd oouuttccoommee
Process measures are easier to
measure than outcome measures,
and can be used to provide
immediate feedback to clinicians
However, it is important to link
process measures with their
counterpart outcome measures
“Outcome measures . . . are
ultimately what patients care
about” (Berenholtz et al., 2002)

11. OOuuttccoommeess
Whilst the individual components of a care
bundle each has a strong evidence base,
there is, as yet, limited evidence that
demonstrates that clustering components in
this way improves patient outcome
(Fulbrook Mooney, 2003)

12. AAllll oorr nnootthhiinngg mmeeaassuurreemmeenntt
PPaarrttiiaall ccoommpplliiaannccee nnoott mmeeaassuurreedd
AAllll eelleemmeennttss mmuusstt bbee ccoommpplliieedd wwiitthh
uunnlleessss mmeeddiiccaallllyy ccoonnttrraa--iinnddiiccaatteedd
Management of early severe sepsis
• USA, UK, New Zealand Australia (n = 2,461)
• Scenario-based questionnaire (MC) given to ED, acute medicine, and critical
care doctors
• Based on 6-hour resuscitation bundle
• Only 2 respondents complied with all SSC guideline recommendations
• 81% identified reasons they could not implement at least one of the
recommendations
Reade et al. Emerg Med J 2010; 27: 110-115.

13. CCoommpplliiaannccee
Component
Overall element compliance
DVT 29/32 = 90.6% SH 9/32 = 28.1% GUP 25/32 = 78.1% HOBE 27/32 = 84.4%
Patient 1 on day 1 = non-compliant = 0%
Patient 1 on day 8 = compliant = 100%
1
e.g. DVT
prophylaxis
2
e.g. sedation
hold
3
e.g. GU
prophylaxis
4
e.g. head of
bed
Patient 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4
Day 1 Y Y Y Y N N N N N N Y Y Y N Y Y
Day 2 Y Y Y Y N N N N N N Y Y Y N Y Y
Day 3 Y N Y Y N N N N Y Y Y Y Y Y Y Y
Day 4 Y Y Y Y Y Y N N Y N Y Y Y Y N N
Day 5 Y N Y Y Y Y N N Y Y Y Y Y Y Y Y
Day 6 Y Y Y N Y N N N N N Y Y Y Y Y Y
Day 7 Y Y Y Y Y N Y N Y Y Y Y Y Y Y N
Day 8 Y Y Y Y Y N Y N Y Y Y Y Y Y Y Y
Care Bundle Compliance. Y = compliance, N = non compliance
Overall
Compliance
9/16 = 56.3%
14/16 = 87.5%
Total =
90/128
= 70.3%
ICU compliance day 1 = 0%
ICU compliance day 8 = 50%
All or Nothing
Compliance

14. Run charts
Care Bundle Component Compliance
150
100
50
0
1 2 3 4 5 6 7 8
Audit Days
Compliance (%)
DVT
prophylaxis
Sedation hold
GU prophylaxis
Bed head
elevation

15. Run chart: daily compliance
Care Bundle Compliance
120
100
(%)
Compliance 80
60
40
20
0
Audit Days Overall compliance
Target compliance
All or nothing compliance
Cumulative T reliability
1 2 3 4 5 6 7 8

16. Compliance and outcome
• Assessing compliance
provides limited information
in itself.
• Run charts can also be used
to demonstrate the impact of
interventions on outcome
measures.
• Assessing compliance
against a local outcome
measure is more likely to
demonstrate the impact of a
new intervention.
• To demonstrate effective
change data would need to
be collected for at least six
months, and preferably a
year.
Audit
feedback
Run chart demonstrating effect of compliance
on an outcome measure

17. Bundle Reliability
Measures of reliability inform teams about the extent
of error-free operation of their clinical processes.
This measure tells teams how often every element
included in the relevant bundle was completed in
each patient during the review period.
Not following every indicated bundle element in each
patient constitutes an error.
Defined as the % of cases for whom all applicable
bundle elements are completed.
The suggested reliability goal for compliance is 95%
– Numerator: The number of relevant cases that completed all
elements of the bundle
– Denominator: The total number of relevant cases

18. RReelliiaabbiilliittyy
Component
1
e.g. DVT
prophylaxis
2
e.g. sedation
hold
3
e.g. GU
prophylaxis
4
e.g. head of
bed
Patient 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4
Day 1 Y Y Y Y N N N N N N Y Y Y N Y Y
Day 2 Y Y Y Y N N N N N N Y Y Y N Y Y
Day 3 Y N Y Y N N N N Y Y Y Y Y Y Y Y
Day 4 Y Y Y Y Y Y N N Y N Y Y Y Y N N
Day 5 Y N Y Y Y Y N N Y Y Y Y Y Y Y Y
Day 6 Y Y Y N Y N N N N N Y Y Y Y Y Y
Day 7 Y Y Y Y Y N Y N Y Y Y Y Y Y Y N
Day 8 Y Y Y Y Y N Y N Y Y Y Y Y Y Y Y
Care Bundle Compliance. Y = compliance, N = non compliance
On days 4 5, only patient 1 was compliant with all 4 elements
On days 7 8, patients 1 and 3 were compliant with all 4 elements
Overall bundle reliability = 6/32 = 18..88%%
If reliability 95% in any month, teams should identify which specific
bundle elements are not being reliably executed.

19. SSuurrvviivviinngg SSeeppssiiss
AA gglloobbaall pprrooggrraamm ttoo::
RReedduuccee mmoorrttaalliittyy rraatteess iinn
sseevveerree sseeppssiiss
• Sepsis Resuscitation Bundle:
Tasks that should begin immediately, but must be done within
6 hours for patients with severe sepsis or septic shock.
• Sepsis Management Bundle:
Tasks that should begin immediately, but must be done within
24 hours for patients with severe sepsis or septic shock.

20. NNuurrssiinngg ccoonnssiiddeerraattiioonnss
AAiittkkeenn eett aall 22001111
DDeellpphhii ssttuuddyy bbaasseedd oonn SSuurrvviivviinngg SSeeppssiiss CCaammppaaiiggnn
gguuiiddeelliinneess
GGRRAADDEE uusseedd ttoo rraattee qquuaalliittyy ooff eevviiddeennccee
6633 rreeccoommmmeennddaattiioonnss ddeessiiggnneedd ttoo ccoommpplleemmeenntt SSSSCC
gguuiiddeelliinneess
RReesseeaarrcchh nneeeeddeedd uurrggeennttllyy
– RReeccooggnniittiioonn ooff ddeetteerriioorraattiioonn//ddiiaaggnnoossiiss ooff sseeppssiiss//eeffffeecctt ooff
eeaarrllyy rreessuusscc mmeeaassuurreess
– MMeetthhooddss ooff hhaaeemmooddyynnaammiicc ssuuppppoorrtt//aasssseessssmmeenntt
– TTyyppee aanndd eeffffeecctt ooff ssuuppppoorrttiivvee ccaarree ee..gg.. pprreessssuurree iinnjjuurryy
pprreevveennttiioonn
– AApppplliiccaattiioonn ttoo ppaaeeddiiaattrriiccss

21. Sepsis bbuunnddlleess:: mmeettaa aannaallyyssiiss
CChhaammbbeerrllaaiinn eett aall 22001111
22000044-1111
2211 sseeppssiiss bbuunnddllee ssttuuddiieess:: ttoottaall nn 2233,,443388
SSuurrvviivvaall bbeenneeffiitt::
– 66hhrr rreessuusscc bbuunnddllee == hhiigghheesstt bbeenneeffiitt:: OORR 22..1122 ((1111
ssttuuddiieess;; nn == 11,,881199))
– 2244hhrr bbuunnddllee == lloowweesstt bbeenneeffiitt:: OORR 11..6655 ((55 ssttuuddiieess;; nn ==
1166,,552211))
– CCoommbbiinneedd bbuunnddllee bbeenneeffiitt:: OORR 11..7744 ((1122 ssttuuddiieess;; nn ==
55,,009988))
– SSccvvOO22 7700%% bbeenneeffiitt:: OORR 22..0055 ((55 ssttuuddiieess;; nn == 1111,,553300))

22. CCoommpplliiaannccee MMoorrttaalliittyy 22001133
Autho
rs
Setting Design n Measures Reported compliance Mortality
Almeida et
al., 2013
Portugal
1 ICU
Prospective
cohort
300 · 6h compliance
· Compliance of each
element
· Day vs night
compliance
· All or nothing compliance 2%
· Element compliance range 4-100%.
· Element compliance better at night
Hospital mortality
· day 40%
· night 34%
Chou et
al., 2013
Taiwan
1 ICU
Before and after
study: 4 phase
(modified
bundle)
55,
30,
26, 53
· 6- and 24h
compliance
· All or nothing for
both bundles
· Before 20% (phase 1)
· After 79.2% (phase 4)
Hospital mortality
· Before 43.6%
· After 10.0%
· (lactate mortality OR 2.2)
Miller et
al., 2013
USA
11 hospitals
18 ICUs
Prospective
cohort (3 phase)
4329 · Combined 6h- and
24h bundle
· “All-or-none
compliance…non-compliance
with any
single element
interpreted as non-compliance
with
bundle”
· Baseline compliance 4.9%
· Post-implementation compliance 73.4%
Hospital mortality
i) Bundle Compliant
· Baseline 21.2%
· Post-implementation 8.7%
ii) Bundle non-compliant
· Baseline 21.7%
· Post-implementation 9.7%
Zhi-qiang
et al., 2013
China
11 hospitals
Prospective
cohort
218 · 6- and 24h
compliance
· Overall compliance:
“when patients met all
relevant targets”
· 6h 5.5% (n = 12)
· 24h 17.4% (n = 38)
· 28 day mortality 33.0%
· 6h comp 25.0%; non-comp 33.5%
· 24h comp 26.3%; non-comp 34.4%

23. 22001122 SSeeppssiiss GGuuiiddeelliinneess
DDeelllliinnggeerr RRPP,, LLeevvyy MMMM,, RRhhooddeess AA,, eett aall..
SSuurrvviivviinngg SSeeppssiiss CCaammppaaiiggnn:: IInntteerrnnaattiioonnaall
gguuiiddeelliinneess ffoorr mmaannaaggeemmeenntt ooff sseevveerree sseeppssiiss
aanndd sseeppttiicc sshhoocckk:: 22001122..
CCrriittiiccaall CCaarree MMeeddiicciinnee.. 22001133 FFeebb;;4411((22))::558800-
663377..

24. SSeeppssiiss bbuunnddllee mmooddiiffiiccaattiioonnss
TThhee ‘‘oolldd’’ SSeeppssiiss RReessuusscciittaattiioonn Buunnddllee
wwaass mmooddiiffiieedd iinnttoo ttwwoo nneeww bbuunnddlleess::
– SSeevveerree SSeeppssiiss 33-HHoouurr RReessuusscciittaattiioonn
Buunnddllee
– 66-HHoouurr SSeeppttiicc SShhoocckk Buunnddllee
TThhee ‘‘oolldd’’ SSeeppssiiss MMaannaaggeemmeenntt Buunnddllee
hhaass bbeeeenn eelliimmiinnaatteedd
– NNeeww:: OOtthheerr SSuuppppoorrttiivvee TThheerraappiieess

25. SSeevveerree SSeeppssiiss 33-hhoouurr
RReessuusscciittaattiioonn Buunnddllee
MMeeaassuurree llaaccttaattee lleevveell
OObbttaaiinn bblloooodd ccuullttuurreess ((bbeeffoorree ggiivviinngg
aannttiibbiioottiiccss))
AAddmmiinniisstteerr bbrrooaadd ssppeeccttrruumm aannttiibbiioottiiccss
AAddmmiinniisstteerr 3300 mmLL//kkgg ccrryyssttaallllllooiidd ffoorr
hhyyppootteennssiioonn oorr llaaccttaattee 44 mmmmooll//LL ((3366
mmgg//ddLL))

26. 66-hhoouurr SSeeppttiicc SShhoocckk Buunnddllee
11.. AAppppllyy vvaassoopprreessssoorrss iiff hhyyppootteennssiivvee ((wwhheenn
nnoonn-rreessppoonnssiivvee ttoo iinniittiiaall fflluuiidd rreessuusscc)) ttoo
mmaaiinnttaaiinn MMAAPP 6655 mmmm HHgg
22.. PPeerrssiisstteenntt hhyyppootteennssiioonn ddeessppiittee vvoolluummee
rreessuusscc oorr iinniittiiaall llaaccttaattee 44 mmmmooll//LL
– MMaaiinnttaaiinn CCVVPP ((ttaarrggeett 88 mmmm HHgg))
– MMaaiinnttaaiinn cceennttrraall vveennoouuss OO22 ssaattuurraattiioonn ((ttaarrggeett
7700%%))
33.. RReemmeeaassuurree llaaccttaattee iiff iinniittiiaall llaaccttaattee eelleevvaatteedd
((ttaarrggeett == nnoorrmmaall;; aarrtteerriiaall 11..66 mmmmooll//LL,,
vveennoouuss 22..22 mmmmooll//LL))

27. Severe SSeeppssiiss Buunnddlleess:: OOtthheerr
SSuuppppoorrttiivvee TThheerraappiieess
11.. Blloooodd pprroodduucctt aaddmmiinniissttrraattiioonn
22.. MMaaiinnttaaiinn aaddeeqquuaattee ggllyyccaaeemmiicc ccoonnttrrooll
33.. MMeecchhaanniiccaall vveennttiillaattiioonn ooff sseeppssiiss-iinndduucceedd
AARRDDSS
44.. SSeeddaattiioonn,, aannaallggeessiiaa,, aanndd nneeuurroommuussccuullaarr
bblloocckkaaddee
55.. DDVVTT aanndd ppeeppttiicc uullcceerr ddiisseeaassee pprroopphhyyllaaxxiiss
66.. NNuuttrriittiioonn
77.. SSeettttiinngg ggooaallss ooff ccaarree

28. 3 Major Sepsis Trials
ProCESS (Protocolized Care
for Early Septic Shock)
– USA
– 3 arms
ARISE (Australian
Resuscitation In Sepsis
Evaluation Randomised
Controlled Trial)
– Australia, commenced 2008
– 2 arms
ProMISe (Protocolised
Management in Sepsis Trial)
– UK
– 2 arms
Adult ED patients
Inclusion criteria based on
Rivers et al (2001). Early
goal-directed therapy in the
treatment of sepsis and
septic shock. New Eng J
Med 345:1368-77.

29. ProCESS
Protocolized Care for Early Septic Shock
5-year RCT study
31 USA emergency departments (n = 1341)
Three approaches to sepsis care
– Early goal directed therapy
Sepsis Resuscitation Bundle
– Protocolized standard care
Does not mandate central lines, inotropes or blood
transfusion
– Usual care

30. ProCESS Outcomes March 2014
ProCESS set out to determine whether a specific
protocol would increase survival in patients with
septic shock.
It showed that patient survival was essentially the
same in all three treatment groups
– 60-day mortalityEGDT (n = 439) 21%
Protocol (n = 446) 18.2%Standard care (n = 456) 18.9%
– 90-day and 1-year mortality ns
indicating that sepsis patients in these clinical
settings were receiving effective care
N Eng J Med (May 2014) 370(18): 1683-93

31. Surviving Sepsis Campaign:
Response to ProCESS
The SSC will determine (any) appropriate revisions to the bundle elements
when the ARISE and ProMISE study results are available.
ProCESS does not address the protocolized management of patients with
severe sepsis without septic shock
The ProCESS results have no impact on the 3-hour bundle
Regarding the SSC 6-hour bundle (2):
– A companion paper appears to support a mean initial arterial pressure (MAP) target
of 65 mm Hg, which is one of the indicators in this bundle. (5)
– The ProCESS paper does not address repeating lactate measures in patients with
elevated lactate while literature supports doing so.
– When measured, the first ScvO2 was 71 ± 13%, which is another of the indication
of the bundle.
– The majority of the patients in the usual care (56.5%) and protocol-based standard
care arms (57.9%) of ProCESS had central lines inserted as part of clinical care.
– The 6-hour bundle currently asks only that CVP be measured and that a venous
blood gas be sent from that line to obtain ScvO2. SSC recognizes that alternate
means of obtaining results exist and will address specific ways of including those
data in future iterations of the quality improvement database.

32. ARISE Outcomes
51 centres in Australia NZ (n = 1600)
– 90-day mortality EGDT 18.6% (n = 796 ) Usual care 18.8% (n = 804 )
– No significant differences in:
Survival timeIn-hospital mortality
Duration of organ support
Hospital LOS
– EGDT group
More vasopressor infusion
More blood transfusions
More dobutamine
Conclusion
– EGDT in ED patients with early septic shock does not reduce all-cause 90-
day mortality
N Eng J Med (Oct 2014) 371(16): 1496-505.

33. ARISE 6h differences (p 0.001)
EGDT Usual care
CVC insertion 714 (90.0%) 494 (61.9%)
Fluid volume 1.9 l 1.7l
Vasopressor infusion 66.6% 57.8%
Red cell transfusion 13.6% 7.0%
Dobutamine 15.4% 2.6%
MAP at 6h* 76.5 mm Hg 75.3 mm Hg
ED LOS 1.4 h 2.0 h
* p = 0.04
Also, between 6-72 hours, more EGDT patients received vasopressor
infusion (58.8% vs 51.5% p = 0.004) and dobutamine (9.5% vs 5.0% p
0.001)

34. ProMISE
ICNARC
56 UK EDs, n = 1260
– 2 arms: EGDT usual care
– Early signs of sepsis septic shock
– Primary outcome 90-day mortality
– Cost effectiveness
Data collection Feb 2011 – July 2014
– Closed 24/7/14 – in follow-up
Awaiting results….

35. CCoonncclluussiioonnss
SSoommee eevviiddeennccee tthhaatt iinnccrreeaasseedd ccoommpplliiaannccee rreedduucceess mmoorrbbiiddiittyy
aanndd mmoorrttaalliittyy
VVaarriioouuss bbuunnddllee mmooddiiffiiccaattiioonnss aanndd iinntteerrpprreettaattiioonn ooff wwhhaatt
ccoonnssttiittuutteess ‘‘ccoommpplliiaannccee’’
BBuunnddllee rreelliiaabbiilliittyy nnoott rreeppoorrtteedd,, bbuutt ggeenneerraallllyy vveerryy llooww iiff
ccaallccuullaatteedd
difficult Inconsistent standards of reporting – makes comparison difficultEEaarrllyy ddaayyss:: nnoo ssttuuddiieess rreeppoorrtteedd uussiinngg ‘‘nneeww’’ sseeppssiiss bbuunnddlleess
EEvviiddeennccee lleeaanniinngg ttoowwaarrddss nnoonn--eeffffeeccttiivveenneessss ooff 6hh EEGGDDTT
33hh sseeppssiiss bbuunnddllee nnoott eevvaalluuaatteedd

36. Dr Bronagh Blackwood UK
Prof Maureen Coombs NZ
A/Prof Sharon Irving USA
Prof Ruth Kleinpell USA
Bronte Martin, Australia
Prof Claire Rickard, Australia
Kathleen Vollman USA
Prof Ged Williams, UAE