1. Explain how the principles of the Toyota company can be applied to anesthesiology and the operating room.

2. Summarize the differences between classical and lean process improvement.

3 Describe typical challenges encountered when using lean process improvement methodology in anesthesiology and operative services.

4. Provide examples of disagreement between clinical experience and clinical trials, and compare human vs analytic decision-making strategies.

5. Discuss unique characteristics of human decision making when risk and uncertainty are involved, and identify potential strategies to improve decision making in ambiguous or uncertain environments.

The Toyota Way
Lynn D. Martin, MD, Professor of Anesthesiology and Pediatrics, University of Washington School of Medicine, and Director, Department of Anesthesiology and Pain Medicine, Children’s Hospital and Regional Medical Center, Seattle, WA

Customer-focused approach to improve performance and quality: Seattle Children’s Hospital (speaker’s institution) in process of fully incorporating Toyota methodology and making it preferred method for changing or improving processes; reasons for change include safety concerns, communication errors or lapses, and medication errors; institution continues to struggle with service deficiencies, predominantly related to access

Toyota company methodology: successful for past 50 yr; cost-effective; provides reliable service; responsive to changing whims of customer; quicker delivery of new products; incorporated throughout organization; high levels of employee engagement and morale; focused on “4 P’s” (philosophy, process, people [ie, partners], and problem-solving); 80% of corporations in United States that have tried Toyota methodology have failed (due to impatience with long-term perspective); focuses on process and eliminating waste; delivers more reliable, higher quality products in shorter time

Improve process: respect, challenge, and educate workers in operating room (OR), intensive care unit (ICU), and emergency department (ED) so they are able to solve problems; examine problems in processes and continually improve and learn from experiences; pursue perfection incrementally over time

Lean thinking: focus on waste, complexity, and variation within system to reduce costs; entire system optimized; information technology (IT) helps achieve long-term goals; substantial improvements made by identifying and removing waste and simplifying processes; quality improvement based on value-stream improvement (ie, entire service line)

5S and visual controls: organize work center to avoid need to search for supplies, equipment, or patients; everything should be readily apparent visually; always focus on patient; make problems visible

After identifying problem: do not “pass it on down the line” “mistake proof” system (challenging for service industry); build in checks to identify mistake; avoid letting it turn into defect when passed down; stop production if problem with quality identified; perform rapid root-cause analysis; go to where problem originated and fix it; avoid replicating problem

Delivery of materials: focus on each patient separately (do not batch work; focusing on several patients at once hides errors); utilize external set-up to improve rapidity of turnover; standardize processes so all providers practice similarly; balance amount of work among team members

Continuous performance improvement (CPI): focus on waste; reduce and eliminate variation whenever possible by standardizing work; improve process; test hypotheses; measure improvement (continuous process); data critical to informed thinking; goal to make improvement (“don’t make it perfect, just make it better”)

CPI in action: first project—involved loading dock; became more organized; able to find needed items quickly; improvement sustained; second wave of improvements—introduced into clinical areas (eg, medication turnaround time) in which no physician leadership involved; goal to have medication delivered to unit in ≤90 min; previously, 65% of medication delivered in ≤90 min; after implementing new process, 95% of medication delivered in ≤90 min; currently 98% of medications delivered in <60 min; implementation in clinical settings with physician involvement—first rapid process improvement project involved documentation; previously, surgeon, anesthesiologist, and nurse required to fill out 18 forms before tonsillectomy and adenoidectomy; at end of 1-wk workshop, number of required forms reduced to 11; 3 mo later (online ordering implemented), 3 more forms eliminated; currently down to 7 forms; key principle of waste involves duplication of work (organization saw ≈90% reduction in duplication of documentation); other results—dramatic increase in staff satisfaction; 81% reduction in total steps from admission; value-added percentage of process increased to 70%; scheduled arrival time before surgery reduced (previously, 2 hr; now, 75 min); family experience survey scores dramatically improved; room turnover times reduced by 36%

Role of leadership: creation of infrastructure —oversaw CPI principles throughout hospital; looked at value stream; created visibility rooms; continued leadership training and increased intensity or frequency of workshops; significantly reduced inventory in induction area and moved materials to improve patient flow; OR that was previously used for storage now used for surgery; more floor and shelf space; significantly reduced total number of steps required to obtain anesthesia supplies and inventory; scheduling process improved; CPI principles implemented throughout hospital—in preoperative clinic, resulted in reduced number of steps, increased value, and significantly reduced patient wait time; significant reductions seen in virtually every area of hospital (eg, number of steps, wait times, staff travel distances); most recent improvements—12 value streams managed; >76 improvement projects completed; families now engaged; 50% reduction in nonoperative time for regional blocks

Lessons learned from trip to Japan: reinforce importance of scientific method; begin by standardizing process, then measure performance, create hypothesis to improve process, test hypothesis, measure results, and determine if hypothesis correct; ensure transparency of data at location where work being performed; goal to enhance quality of care for patient and family, but recognize importance of making it easier for those providing care

Interim results: infection rates down; family experience survey scores up; standardized hand-off reports from nurse to nurse, physician to physician; families participating on rounds; faculty and staff engagements at all-time highs; decline from previous year in cost per patient-day

Current situation at Seattle Children’s Hospital: moved from proof-of-concept approach; now focusing on value stream to process improvement; working with suppliers, payors, and referring physicians as part of ultimate lean enterprise; looking forward—3 new value streams added in 2008; leadership training now taught by hospital staff instead of external consultants; point improvements continue to be evaluated, with focus on sustaining successes; developing abilities to cross-audit; more Japanese study trips; challenges include ability to access reliable data, understand work methods, and handle new problems; infrastructure important to implementation of CPI; suggestions— create standard tools and methods and “don’t waver from them”; use information and facts (not innuendo) to define work; track cost savings

Evidence vs Experience: Making Better Decisions
Avery Tung, MD, Associate Professor, Department of Anesthesia and Critical Care, University of Chicago Pritzker School of Medicine, Chicago, IL

Introduction: evaluation of Echinacea vs placebo in experimental rhinovirus infections found no difference in volume of nasal secretions, virus titers, or interleukin levels; however, survey of Echinacea users finds continued belief in efficacy for prophylaxis; speaker acknowledges he also does things each day that are against published evidence (eg, albumin for critically ill; use of pulse oximetry; use of pulmonary artery [PA] catheterization for hemodynamic monitoring)

Pulmonary artery catheter: Connors (1996) looked at 5700 critically ill patients in prospective cohort-match study; found PA catheter associated with more deaths, longer length of stay, and more money spent, compared to those without PA catheter; study results “made people furious”; PA catheter had been in use for ≈35 yr when study released; only rational conclusion that “doctors just use stuff before benefits and hazards are fully evaluated”; critical care providers say, “no, that is not true,” and refute idea that they are using PA catheter because of manufacturer’s influence; nonetheless, finding that PA catheter does not improve outcome “one of the most consistent results” found in medical literature

Can so many physicians all be wrong? studies continue because “we simply can’t believe the result”; entire articles based on trying to bridge gap between experience and literature regarding PA catheter; common explanation that “we just haven’t done the right study yet” (eg, American Society of Anesthesiologists’ [ASA] practice guidelines in 2003); other common criticisms include quality of data (does not “measure what it’s supposed to measure”) and lack of consensus regarding appropriate response to data from PA catheter; could argue monitoring does not improve outcome (eg, critical to practice, yet study from 1994 in 20,000 patients found no perioperative benefit); “training wheel” argument—PA catheter teaches provider how to manage hemodynamics, and once management has been learned, PA catheter becomes unnecessary; speaker refutes based on fact that trauma and cardiac surgeons who use PA catheter most do not discontinue, but continue to use; “sticks in the mud” argument—argued that physicians do not let go of things that are outmoded and refuse to change; eg, use of perioperative β-blockade has not increased beyond 50%; however, speaker disagrees, citing instances of therapies disappearing as result of negative studies (eg, Mangano study of use of aprotinin)

Do humans make decisions rationally? decision making conscious product of human reason; involves identifying possible choices, computing probabilities of specific outcomes from each choice, computing utility that each choice offers, and choosing that which gives best utility; however, “people just don’t do that”; in real world—all available choices unknown; assessment of risk incomplete; immense computation required to calculate likelihood; real world constantly changing (therefore, risk and uncertainty changing); sometimes necessary to act and make decision immediately; strategies used to avoid problems with applying standard rational theory to real world—1) satisfaction with adequate (vs optimal) solution, 2) analytic and intuitive cognitive systems working together, and 3) pattern- matching situations to mental model; testable hypotheses (to show decision making often intuitive)—quality of decisions expected to decrease if time allowed to make decisions decreases; information processing should occur automatically; cognitive illusions expected; study shows master chess players make intuitive (not analytic) decisions; brain processes information on subconscious level (eg, seating location during last visit to favorite restaurant); cognitive illusions plague decision making

Framing bias: judgments influenced by how decision worded; eg, if given choice between A) “give you $240,” or B) “give you 25% chance to win $1000 but 75% chance of winning nothing,” most would choose choice A; however, if given $1000 and choice between A) “take away $740,” or B) “give you 75% chance of losing it all, but 25% chance of losing nothing,” most would choose choice B; thus, therapeutic choices framed in terms of survival elicit different response from choices framed in terms of dying

Representativeness: diagnosis based on how well description matches mental model, rather than basing diagnosis on likelihood of occurrence; eg, likelihood of major flood somewhere in North America in which >1000 people drown vs likelihood of earthquake in California causing flood in which >1000 people drown; specific plausible scenarios seem more frequent than actually occurs

Base-rate neglect: instead of calculating pretest probability of occurrence, occurrence matched to mental model; more likely to happen if it matches mental model; example—Steve described as shy, withdrawn, helpful, and with little interest in people or world of reality; most would consider Steve more likely librarian than salesman, even though more men salesmen than librarians

Ambiguity aversion: most people dislike not knowing likelihood and risks; element of human decision making first described by Daniel Ellsberg in 1961; proposed scenario—consider urn containing 9 balls; 3 are red, other 6 are either black or yellow; ball selected at random from urn; what is color of that ball? most guess red; however, if asked to guess color that ball is not, most guess red based on calculated probability; human preference for certainty in first scenario forces one to select higher probability color; same preference for certainty occurs in medicine

Hindsight: tendency to judge things differently once outcome known; once something happens, everything that preceded occurrence becomes clear; brain attempts to make sense of events once something happens; example— Joe developed cardiogenic shock after his heart attack and died; transition to statement that Joe died of cardiogenic shock after his heart attack links events together; brain has automatically tried to make sense of occurrence

Evidence-based medicine: much of literature actually incorrect; randomized controlled trials no guarantee; because world changing, studies may need to be repeated because results may be different (eg, incidence of peanut allergy [nearly doubled in <5 yr]; prevalence of heparin-induced thrombocytopenia in cardiothoracic ICU [nearly doubling in <10 yr]; tidal volumes; perioperative β-blocker therapy [evidence indicates routine use does not improve outcome as necessary]); systematic reviews outdated at 5 yr

Value of intuitive decision making: humans superior at visual identification; ability to match possible occurrences also better than with algorithm-based approach

How can we do better? acting in statistically incorrect manner best suited for instances when 1) risk uncertain, 2) understanding human behavior important, and 3) implied odds outweigh risk

Making better decisions: train intuition with experience (seek as much experience as possible); understand vulnerability to cognitive illusions (prone to misjudge probability by attempting to match to mental model); recognize limitations of literature; current methods may not accurately assess value of diagnostic or therapeutic interventions; recognize when judgment valuable