What is new in safety thinking and its implications for health care

1
What is new in safety thinking and its
implications for health care

• Carol Haraden, IHI
• René Amalberti, HAS

2
Outline of the Session

• Welcome, introduction of speakers and session
• What is the problem ? (Carol H, 5)
• Two general avenues for making healthcare safer
• Four basic failures mechanisms when designing
  safety rules and safety policies (RA, 10)
• How to address the growing issue of non
  compliance (Carol H, 10)
• Six principles to improve design (Rene A, 10)
• Accompanying solutions to manage deviances (Carol
  H, 10)
• Test yourself on a concrete example (CH RA,
  20),
• Conclusions (Carol H, 5)

3
The problem

• Patient Safety is a priority- what would that
  look like?
• Top of agenda at all levels
• Deep knowledge of safety issues
• Managing risk is critical for everyone (huddles,
  FMEA, handovers)
• Efforts are considerable
• Once the horse has left the barn Root cause
  analysis, media management, law suits, policy
  review etc. etc.
• Fixing the barn door The Hospital at Night
  Programme

Int'l Forum
3
4
A New Set of Problems Change is unavoidable by
design or default

• Service Commitment taking precedence over
  Training and Education
• Night working has little educational value
• Poor continuity of care
• Loss of consultant team structure
• Reduction in medical staff available during
  office hours

5
Key elements of the strategy I

• Minimise workload at night
• Doing things differently
• Drawing work into day
• Effective demand management e.g. through
  primary care out of hours
• Service reconfiguration for some specialist
  services - supported by effective treat and
  transfer arrangements along agreed pathways of
  care

6
Key elements of the strategy II

• Minimise medical workload at night
• work within a multi-disciplinary, competency
  based team
• up skill ward staff to minimise reliance upon the
  night team
• reduce duplication
• take away inappropriate tasks
• effective bleep/call policies
• better use of new technologies
• mobile phones not bleeps
• digital imaging
• e-prescribing
• electronic records

7
Hospital at Night requires generic competencies
Complex procedures RS Intubation External
Pacing Epidural Arterial Blood gases Chest drain,
Suturing Central lines, BiPAP
Specific clinical competencies
Generic Non Clinical Time management Delegation
Risk management Leadership
Generic Clinical Recognition of the
sick Resuscitation ALS Clerking in,
Discharge Pain management, Prescribing Ordering
x-rays, Debriefing, Teaching,
Routine Procedures Venepuncture Blood
transfusion Urinary Catheters /-
Suprapubic Recognition of Death NG tube
8
The problem cont.

• Results (outcomes) are not enough
• Lack of reliable design of processes that drive
  outcomes
• Often no measure, or wrong measure of effective
  actions i.e all-or- nothing measure
• Limited staff adherence

• Why?
• Brownian activity, No enough whole vision
• Design of safety policies and recommendations put
  to the question

Int'l Forum
8
9
The 2-4-6-4 strategy to improve safety policies

• 2 goals to achieve
• 4 generic failures to avoid
• 6 principles to adopt for design
• 4 principles to adopt for the control of
  implementation

10
2. Patient safety Two goals
The safest technical care to reducing errors No
errors in the implementation of the chosen
strategy
Adverse events
Safe outcome (No Adverse Outcome)
Maximizing recovery Standardizing Reducing losses
of opportunity Choice of the best proven medical
decision strategy Evidence-Based Medicine EBM
Standardizing practices (Changement dordre de
présentation)
Int'l Forum
10
11
3- Four basic failures mechanisms when designing
safety policies
12
The Conflict paradigm

• Principle Let hyper specialists optimize alone
  (in a silo) the best safety policy
• May result in conflicts with other safety
  policies / best practices
• Example Use of neuro-muscular blocking agents
  in case of surgery on full stomach

BEST EXPERTS
Silo
Silo of Technical vision
Conflict generation for people hands on
13
The Floor lamp strategy

• Principle Design safety policy and performance
  indicator for monitoring effectiveness
• Problems occur with the choice of
• of performance indicator
• Example
• Use of Hydro-alcoholic solutions for hand
  washing in geriatrics
• Indicator on mean vs indicator on end
• Safe and effective transfusion

14
The Tuesday paradigm

• Design Principle Staffs highest bid (best
  effort) in thinking safety
• Design ideal policy based of best conditions,
  full staff, best competences (the Tuesday
  morning when all staff is present).
• Examples
• Pain management with a permanent infusion of
  anaesthetic drugs using a crural cath.
• Nice technique, obvious benefits for patients
• However still many deaths/ year wrong
  catheter/wrong infusion, Ischemic syndrome (no
  longer pain)
• Why?
• Solution not robust enough in face of periodical
  short staffing conditions
• New cleaning protocols for endoscope tools
  (prion-resistant)
• Solution not robust enough to resist work
  pressure
• Washing hand protocols before 2001 and the
  generalization of hydro-alcoholic solutions
• Solution not robust enough to resist work
  pressure

15
The Politics of small steps

• Principle Small is better than nothing
• Deconstruct problem and take action ANYWHERE
  feasible, ANYSIZE feasible
• Make sense, rational, give impression to improve
• Whats wrong?
• Small steps have small effects, do not convince
  professionals

Barriers safe guards
AEs
Mitigation
Recovery
Prevention

• EXAMPLE wrong patients
• Deconstruct why occur ? Same name, chart
  error, Ask for systematic formal check and
  signature of supervisors before exam (chief
  nurse, or other)
• Problem fixed? No, very low compliance,
  Immediate Ice melting effect,
• And even worse Increase global complexity,
  system become cryptic

Initial system
16
Usual Consequences of Poor Design

• Non compliance
• Deviance
• Even worse paradoxical side effects and safety
  degradation

Carol
17

• 3. Managing the consequences of poor safety
  policy-design

18
The Medication Dispensing Machine

• Med-Dispense delivers an affordable, robust
  system that even the smaller hospital can afford.
  Our systems use the latest interface technology,
  ensuring revenue capture as well as JCAHO
  compliance at a surprising low cost.
• our systems are undoubtedly the most
  affordable as well as the simplest to use.  
• allow nursing access to medications with
  pharmacy control via centralized server reduce
  medication errors eliminate missing doses
  eliminate manual charge functions with interface
• functionality

19

20
Use of the Machine

• Typically, 2 machines for each 13 room hallway
• Nurse has 5 patients placed throughout the unit
• On average, each patient has between 7-10
  medications 2-4 times per day
• Nurse takes medications out 1 at a time

21
System Migration to Unsafe Practices to

All patients meds placed in pockets
Illegal-Illegal space
The guidelines and policy- take meds out for one
pt. at a time
VERY UNSAFE SPACE
Perceived Vulnerability
ACCIDENT

PERFORMANCE
22
Why the migration to less safe practices?

• Policy unmanageable
• Nurses did not have the time to make several
  trips back and forth to the machine several times
  a day
• The medication cart had been taken away
• No one had studied medication administration
  patterns before installation
• Safe dispensing ? safe administration

23
Concept of Border-line Tolerated Conditions of
Use (BTCU)

• The BTCU becomes the stabilized usual level of
  performance
• We do them regularly with only rare adverse
  outcomes though many near hits.
• We come to feel safer and safer, we come to see
  the BTCU as normal and safe.
• Risks are known and supposedly under control.
• Practices are rarely penalized even though they
  are known.

24
The Result of Migration is Well Known (1)

• The result of migrations is a large range of
  illegal practices... which over time became
  normal for everyone, part of the systems
  normal operation.
• Since these practices are illegal, nothing can be
  written about them, to comment or to accept their
  existence.
• The only words written about these are
  ineffective memos reminding the staff about the
  old, written rule. These practices are only
  commented upon verbally.

25
The Result of Migration is Well Known (2)

• There is great reluctance to monitor these new
  practices with indicators since they are not
  supposed to exist!
• It is essential to remember that all stakeholders
  in the system migrate and deviate from standards,
  including management.
• Over time, these practices become a shared
  professional consensus.

26
How to improve

• Improve safety policy design
• Foresee problems
• Adopt a whole vision
• FMEA, other prospective analyses
• Cope with unexpected effects
• Control deviances
• Document lessons learned

Int'l Forum
26
27
René Amalberti

• 4. Six principles for a better cost-effective
  design of safety policies

28
Summary of problems

• (low) perceived benefits
• No or hypothetic experience of negative outcomes
  associated with the policy
• No or little personnel benefits (including
  system rewards)
• (High) burden in implementation
• Slows down the pace of work
• Involves multiple actors (co-ordination), or
  even worse, imposes (hypothetical) additional
  staff
• May impose sacrificial decision on competitive
  tasks / best procedures
• Can impact new or old procedures
• Perceived as an OR procedure Vs AND
  procedure

29
Six principles for a better design

• P1 The Severity and frequency of expected
  outcome
• P2 The System tolerance to non compliance
• P3 The time needed to comply with the policy
  and the total time needed to achieve the work
• P4 The extra-resource needed to comply with the
  policy
• P5 The potential for by-passing strategies to
  make the job effective when the policy is not
  applicable
• P6 The conflict resolution when the policy is
  in competition with others contradictory policies

30
P1 Severity and frequency of expected outcome

• Test the frequency of the targeted outcome in
  your ward
• If none in past years or if the scale of effect
  small
• Bad prognosis
• Lesson
• If low frequency or little effect, then
  continuous reinforcement program needed
• Need staffing on Adverse outcomerelated case
  based stories
• Need long term managerial commitment to maintain
  compliance
• Severity of effect
• Change behavior for loosely related multiple
  outcomes (e.g. make more checks, refuse
  interruptions, reduce distractions) will be
  sacrificed easily
• Adopt active patterns for a clear and unique
  patient outcome associated with a high frequency
  risk of AE (post surgery phlebitis prevention)

31
P2. System tolerance to non compliance

• Suppose you - or any member of the staff - are
  not complying, what occurs?
• No change, no impact forget the policy
• Who will say something?
• Managerial reaction ? Test it with realistic
  values

32
P3 Time needed to comply with the policy and
total time needed to achieve the work

• Consider the TOTAL time needed to achieve the
  work sequence where the new policy is active
• Calculate the maximum work feasible
• Infer sacrifices if the demand exceeds the time
  available
• Lesson
• DO not add
• REPLACE

33
P4 Extra-resource needed to comply with the
policy

• Consider the worse usual situation during the
  week, with the less competent and numerous staff
• Is the new policy still applicable
• If not, forget it, or design two a policy for
  best circumstances and a policy B compatible with
  degraded situation

34
P5 Potential for by-passing strategies when the
policy not applicable

• Imagine that you cannot apply the policy
• How will you do your job?
• Do as you did before
• Do different (old time no longer feasible)
  Brainstorm how?
• Imagine the worse, you will be behind real!
• Can you block these potentially dramatic
  scenarios?
• Imagine side effects and the cost of control
• If not affordable, forget the procedure

35
P6 Usual conflict resolution when the policy is
in competition with others contradictory policies

• Staff and test with end-users potential
  conflicting strategies
• If potential conflict, foresee what will be the
  usual sacrifices.
• Adopt a whole vision
• Address conflict resolution

36
Trading function of Good Vs Bad design
Safety policy Induced AEs !!
Perception of benefits
Require efforts (ask for sacrifice) Potential sid
e effects
Comprehensive and intuitive reduction of adverse
outcome
Easy to do, every time (resources and time
available)
Growing Non Compliance
Why the safety policy not understood or
misunderstood
Procedures not achievable
Objective safety benefits
Converse effects greater than expected benefits
Loss of efficacy
37
Summary score your chance of success
38

• 5. Four steps for a better control of migrations
  and deviance

39
Detecting Becoming aware of the safety problem
and sharing what is happening

• There is no longer a medication cart or a way to
  separate individual patients meds and carry to
  patient bedside
• Clear that medications taken out of machine and
  put in pockets is unsafe
• When and where can we have this conversation?

Int'l Forum
39
40
Controlling Team awareness of violation and
steps to prevent escalation of unsafe practice

• The care team agrees that their practice is a
  significant deviation and a safety problem
• They agree on
• a new system, and
• a new set of operating rules.

Int'l Forum
40
41
Blocking Prevention of future violations

• Transparent and understood consequences for
  violations from agreed upon system
• Swift and clear action taken when violations occur

Int'l Forum
41
42
Accepting adjust and change the safe space to
accommodate new, safer practice

• They re-write the policy describing the new
  procedure. They agree to
• measurement of the effectiveness and safety of
  the new system to detect problems, and
• supporting each other by becoming accountable for
  the safety of the system.
• Discuss with risk management

Int'l Forum
42
43
Questions, clarifications before examples?
44
ExamplesYou see a problemYou see a team making
safety decisions in reactionPlease react,
identify potential drawbacks, and suggest
improvements
45
Case 1 Intrathecal vincristine
Partially after Croskerry, 2005, Intrathecal
vincristine, an archetype among medical errors,
37th conf Pediatrics oncology

• Context treatment of leukemias and lymphomas
• A vinca alkaloid (usually vincristine) is
  prepared for administration
• Methotrexate is prepared for intrathecal
  administration to eradicate malignant cells
  sequestered within the blood-brain barrier
• The two medications are erroneously switched,
  vincristine is given intrathecally (highly
  neurotoxic)
• No effective recovery strategy, Majority of cases
  die within days
• Example
• Patient needing cure, decide to move the day of
  RDV for social reasons
• No beds available at oncology outpatient clinic
  that day,
• Patient moves temporary to a medical ward
• Pharmacy deviates from standard protocol for
  drug delivery, send packages of vincristine and
  methotrexate
• Local resident poorly trained, ask assistance
  from oncology department. They ask another
  resident from another ward (former resident of
  the oncology department) to help
• Was noon, lunch time for staff
• Error

46
Intrathecal vincristine (continue)

• What to do?
• No more vincristine delivered outside outpatient
  clinic ?
• Mandatory presence of a senior oncologist anytime
  intrathecal not administrated in outpatient
  clinic ?
• Test the SIX DESIGN principles for each policy

47
Summary score your chance of success
48
Case 2 Medications in Labor and Delivery

• Context Operating room on the labor and delivery
  unit
• Fully functioning operating room set up for both
  planned and emergent operative deliveries
• Medications stored by name in small open boxes
  which are labeled with medication name
• After a emergent and difficult C-section, the
  baby is delivered and is doing well after
  stimulation and suctioning
• The Mom is bleeding heavily
• Vitamin K is ordered for baby the nurse reached
  into the vitamin K box, draws up the medication
  and discovers that she has administered methagin
• Example
• Patient and husband were non-English speaking
  and interpreter was late
• Patient could not communicate pain and pressure
• Extremely busy labor and delivery with several
  challenging cases
• Staff had no breaks or evening meal
• Three handoffs as more experienced nurses were
  sent elsewhere

Int'l Forum
48
49
Medications in Labor and Delivery (continue)

• What to do?
• -punish nurse?
• -lock medications up in cart?
• -double check of all medications?

Int'l Forum
49
50
Summary score your chance of success
51
Take-Home Points

• Safety and risk reduction a priority
• No change in outcomes without reliable and tested
  design by end users
• Test the design against the six principles
• MEASURE effectiveness
• Expect migrations and plan for them
• Impact on Design and Training programs

Int'l Forum
51