17-‐07-‐15
1
Andreas Voss Andreas Widmer
Sbarbaro Clin Infect Dis 2001,33:S240
Changing physicians behavior is considered by many to be an exercise in fuHlity – an unaJainable goal intended only to produce premature aging in those seeking the change. The more opHmisHc might describe the process as uniquely challenging.
= coordinated set of acHviHes designed to change specified behavior paJern
… and that you are Dutch
Lets assume that you are interested in hand hygiene …
By far the lowest
observed %
published
compliance with
HH
¤ 24 hospitals in the NL (ICUs & surgery)
¤ 19% overall compliance
² 3% before ² 22% a]er
¤ No difference between RN en MD
¤ Smaller hospitals beJer than bigger
Vicky Erasmus Inf Control Hosp Epidemiol 2009 The Lowlands (NL)
21%
The Honestlands
50-‐70%
The Lielands
80-‐95%
The Rest of the World
17-‐07-‐15
2
hJp://www.hha.org.au/LatestNaHonalData.aspx
¤ We look, others don’t … ¤ Our surveillance is too good (no Hawthorne effect) …
¤ No one can do beJer … ¤ Five indicaHons are (two) too many …
¤ Handhygiene is only one of the measures …
¤ EducaHon ¤ IncenHves for appropriate behaviour ¤ Penalise inappropriate behaviour ¤ …. ¤ No target behavior analysis ¤ No theory on predicted mechanism of acHon ¤ Not considering impulsivity, habit, self-‐control, associaHve learning, …
Standard approach
¤ First canvas the full range of opHons available ¤ Second raHonal selecHon from them
² to match the behavioral target, the target populaHon and the context in which in which intervenHon will be delivered
Improvedapproach
Michie et al. ImplementaHon Science 2011;6:42
17-‐07-‐15
3
Individual’s psychological & physical capacity to engage in the acHvity concerned. Having knowledge & skills
Michie et al. ImplementaHon Science 2011;6:42
Brain processes that direct behavior:
conscious decision making, emoHnal responding, habits
Factors that lie outside the individual that make behavior possible or
prompt it
``````
```````````
This model of behavior allows designing interven4ons
Michie et al. ImplementaHon Science 2011, 6:42
Michie et al. ImplementaHon Science 2011, 6:42
Michie et al. ImplementaHon Science 2011, 6:42
17-‐07-‐15
4
A wise man that you shouldn't mix it in the fruit salad.
A smart man knows that a tomato is a fruit.
… but the fact that I don’t like tomato in my fruit salad doesn’t mean that I am wise!
“Diffusion is the process in which an innovaHon is communicated through
certain channels over Hme among the members of a
social system”
Slides on diffusionin part from Sara Gossgrove
¤ Rela4ve advantage ² Is it beJer than what we already have according to the users (not you)?
¤ Compa4bility ² Is it consistent with exisHng values and pracHces?
¤ Simplicity/ease of use ² Is it easy to understand/adopt?
¤ Trialability ² Can the innovaHon/change be experimented with on a limited basis?
¤ Observability ² Can users see the results of the innovaHon/change?
¤ Categories of adopters ² Innovators (2.5%) ² Early adopters (13.5%) ² Early majority (34%) ² Late majority (34%) ² Laggards (16%)
The diffusion curve generally takes off a]er 10-‐20% adopHon
Not for HH
¤ PercepHon Failures CommunicaHon Failures ¤ Engagement Failures ¤ LogisHcal Failures
17-‐07-‐15
5
¤ Those non-‐compliant don’t believe the evidence or think their situaHon is different
¤ What can you do? ² Repeated conversaHons
² Face-‐to-‐face preferred ² Show data that non-‐compliance harms paHents Address message to all levels of physicians (trainees to clinical leadership)
² Show that your recommendaHons are similar to those from other similar insHtuHons (peer-‐pressure)
¤ RecommendaHons are too complex or we sound doubrul
¤ What can you do? ² Make a single recommendaHon or provide no more than two opHons ² Avoid decision paralysis
² Be confident ² Be enthusiasHc
decision paralysis
¤ Don’t care about hand hygiene ² Involve teams in making local guidelines
² Feedback observaHons and provide posiHve reinforcement
² Appeal to emoHons ² Cases of real paHents who were harmed
² Evidence that the individual paHent can be harmed
¤ Missing means to disinfect hands or reinforce compliance ² sufficient dispensers and availability at point-‐of-‐care
² Talking walls/reminders
¤ Missing feed-‐back ¤ Not involving KOL’s
¤ While we wish to make the most thoughrul, fully considered decision possible … we frequently resort to comply on basis of a single piece of informaHon (trigger)
¤ “Click-‐whirr” (automaHc response, fixed acHon paJern)
17-‐07-‐15
6
¤ Commitments
¤ OpportuniHes for reciprocaHon
¤ Compliant behavior of similar others
¤ Feelings of liking & friendship
¤ Authority direcHves
¤ Scarcity informaHon
¤ Shortcuts – judgmental heurisHcs
¤ Contrast principal
¤ Consistency
Shortcuts-‐ contrast -‐ consistency
¤ “Expensive = good” (and opposite) ¤ “If an expert said so, it must be true”
² Airline industry “captainiHs”
Shortcuts-‐ contrast -‐ consistency
¤ Perceptual contrast
² Buying suite and sweater à expensive item first
² Selling the car first, than the opHons
Shortcuts-‐ contrast -‐ consistency
Shortcuts-‐ contrast -‐ consistency
¤ Most people have a strong desire to look consistent within their words, beliefs, autudes
¤ Whenever one takes a stand that is visible to others, there arises a drive to maintain that stand in order to look like a consistent person.
¤ Therefore, wriJen-‐down and publicly made commitments can be used to influence others and ourselves
Shortcuts-‐ contrast -‐ consistency
17-‐07-‐15
7
¤ At the beach
² Leaving a blanket with valuables – pretended the] – 4/20 people in the area reacted
² Saying “watch my things” first à 19/20 reacted
If I can get someone to commit (take a stand or go on record) I can set the stage for automaHc consistency !
¤ QuesHon to home owners – place large billboard “Drive Carefully” in your lawn
¤ Two groups: A) 17% complied B) 76% complied
¤ Why the difference between A and B ? B was asked a few weeks before to display a small 3x3in sign in window “Be a safe driver”
Shortcuts-‐ contrast -‐ consistency
¤ Key-‐person as role-‐models ¤ Published list of supporters
² WHO pledge
¤ Pre-‐prime your customer ² Would you (theoreHcally) be supporHng our goals? Followed by a visit to commit them to IC
Shortcuts-‐ contrast -‐ consistency
www.sciencexpress.org / 20 November 2008 / Page 1 / 10.1126/science.1161405
§ Orderly alley: 33% liJering § GraffiH alley: § 69% liJering
¤ when people observe inappropriate behavior, this weakens their concern for appropriateness
§ Orderly alley § No graffiH sign § Flyer on handlebar § avoid liJering
We need “order” in our hospitals and people showing appropriate behavior !
17-‐07-‐15
8
¤ “we are obligated to future repayment”
¤ Regan et al. “Coke experiment”
² Joe gave others (unasked for) a drink. Later asked them to buy raffle Hckets from him à those who got a Coke bought more Hckets à those who “liked” Joe bought more Hckets
² but …. relaHonship between “liking” and compliance wiped out in those that received a Coke !
“a feeling of indebtedness by doing us an uninvited favor” ReciprocaHon
¤ Mailing a $5 gi] along with a survey was twice as effecHve as a $50 payment for sending it back (James & Bolstein 1992)
¤ à a small iniHal favor can produce a sense of obligaHon to agree to a larger return favor, since most of us find it highly disagreeable to be in a state of obligaHon
ReciprocaHon
¤ Reciprocal concessions
² Boy scout selling $5 Hckets for Saturday night event, when rejected offering $0,50 candies (buyer feels that second request is a concession to him)
¤ Rejec4on-‐then-‐retreat strategy
² Over ask first, than come with intended amount
² Ask volunteers for 1 week of community service than ask for 1 hour
ReciprocaHon
A nice Christmas gi] for your HCWs –
shortly before launching
your HH campaign
ReciprocaHon
¤ Principle of Social Proof ² Important means that people use to decide what to believe or how to act is to look what other believe or do
¤ Social proof is most influenHal under two condiHons ² Uncertainty ² Similarity
Similar others/social proof
17-‐07-‐15
9
¤ “We view behavior as correct to the degree we see other performing it” ² When lots of people do it – it must be right
¤ Social proof is most influenHal under two condiHons ² Uncertainty ² Similarity
¤ Similar others/social proof
¤ Laughing tracks ² Others laugh = it must be funny
² we react to the sound (even if fake) – not the content
² works even if the sound is arHficial ¤ Church usher salHng collecHon baskets ¤ InvenHon of shopping carts (S. Goldman 1934)
² Shoppers stopped shopping when basket full ² First shopping carts not used ² Hire fake shoppers using carts throughout the shop ?
Similar others/social proof
Similar others/social proof Asch conformity experiments
Uncertainty Heart aJack or drunk? Other bystanders reduce personal responsibility
Similar others/social proof
17-‐07-‐15
10
¤ Never again start a lecture with only 40% disinfect their hands … ² “all HCW agree that HH is the most important …”
Similar others/social proof
¤ “ We most prefer to say YES to the request of people we know and like”
¤ Factors leading to liking: ² physical aJracHveness (halo effect) ² similarity ² familiarity ² praise ² associaHon
Liking
¤ Occurs when a one posiHve characterisHc of a person dominates the way a person is viewed
² Good-‐looking = talent, kind, honest, intelligent
² Works in many situaHons: even judicial process à handsome men lighter sentences
Liking
¤ “We like people that are similar to us” ² more likely to help those that dress like us
Liking
Familiarity
• “Increases through repeated contact under posiHve circumstances”
• “We are phenomenal suckers for flaJery” – PosiHve comments produce just as much liking for the flaJerer independent if they were untrue or true
• “Teamspirit” – Car salesperson baJling with his boss to give us a good price
– Good Cop/Bad Cop
Liking
¤ The nature of the bad news infects the teller even when the person did not cause the bad news ² Blame the weatherman
Liking
17-‐07-‐15
11
¤ An innocent associaHon with either bad or good things will influence how people feel about us or a product ² The good looking models next to the car ² RaHng of idenHcal car changed with and without model
² Men didn’t believe that their judgment was influenced
Liking
¤ Pavlov ² Normal response to food: salivaHon, good and favorable feeling ² AJach that feeling to any goal
² “luncheon technique” (ask during or a]er meal) ² Subjects become fonder of people and things they experienced while eaHng
Liking
¤ Praise instead of criHcize ! ¤ Let others report about the outbreak & extra measures
¤ InfecHon Control luncheon
Liking (aJracHveness, praise, familiarity, similarity, associaHon)
¤ Air freshener ² Will be using it in cars of nursing home MD’s as part of an HH improvement project
Liking (aJracHveness, praise, familiarity, similarity, associaHon)
Would love to have them at doors to
paHent’s room
¤ Power of authority (Milgram study, 1974)
¤ Symbols of authority ² Titles ² Clothing ² Automobiles
Authority
17-‐07-‐15
12
Authority
Expert/Authority
¤ Various types of HCWs understanding their level of the job
¤ No one may overrule a doctor’s judgment à automaHc obedience to doctors’ order ² MD order for paHents with earache: “place in R ear”
² Neither nurse no paHent quesHon the rectal treatment of the earache
¤ People don’t consider the situaHon as a whole but aJend or respond to only one aspect of it
Authority
¤ TV commercial featuring Robert Young (not a medical authority) warning people against the danger of caffeine and recommending caffeine free coffee
¤ RY is associated in the mind of the US public as trusted Marcus Welby, MD from a long-‐running TV series
Authority
¤ Appearance of authority is enough to get is into the click,whirr mode
¤ Symbols of authority ² Titles ² Clothes ² Trappings
Authority
¤ PresHgious Htles lead to height distorHon ² Visitor to college presented as “Student” to “Professor” at 5 different levels, college students than asked to esHmate length: Δ 0.5 per category or 2.5 in overall
² Size and status is related
Authority Germany Netherlands
How high? No! Why? Yes … but I stay put
Authority (Voss, not in book)
17-‐07-‐15
13
¤ Motorists wait significantly longer before honking their horns when stopped in front of a green light behind a luxury car versus an older economy model
¤ True for all symbols of authority: people grossly underesHmate the effect of authority influence on themselves
Authority
¤ People assign more value to opportuniHes when they are less available ² “limited number” and “deadline”
² Difficult to get = valuable
² Loosing freedom of choiceà increases wish
¤ Psychological reactance
¤ Scarcity and informaHon
scarcity, censorship, reactance
¤ Announce a limitaHon of hand-‐rub and than find a “new source of supply”
Authority (obedience, Htle,clothing, cars)
¤ “Things that are difficult to get are beJer” ¤ When free choice is limited or threatened the need to retain our freedom makes us want them significantly more ² “Terrible two and equally desirable toys of which one behind a Plexiglas barrier
² Parental interference with young love
scarcity, censorship, reactance
¤ Response to “banned” informaHon ² Want the informaHon even more ² Believe the informaHon more even though it wasn’t received
scarcity, censorship, reactance
¤ University undergrads’ wish to read a novel a]er 2 different ways of presentaHon …
Restricted t
o those of
21 years and
older
Zellinger 1974
17-‐07-‐15
14
¤ Taste & rate cookies from jar ² jar with 10 cookies ² jar with 2 cookies (scarcity -‐> rated more favorably)
scarcity, censorship, reactance scarcity, censorship, reactance
¤ University HCWs’ wish to read CDC IC guidelines a]er 2 different ways of presentaHon …
Restricted t
o those
highly educa
ted
Limited Edition
Exclusive for HCWs
Where you wondering why the handouts are
so scarce?
¤ IntroducHon of new hand-‐rub or catheters in your hospital ² Present & announce, restrict it to ICU and those units that first order
scarcity, censorship, reactance