Welcome to the Leadership for Safety Webinar Reliability: Keeping Our Promises The webinar will be starting momentarily… If you are having technical difficulties please contact 202-495-3356 or ltiscornia@naph.org Chat Box Please use the Chat Box on the webinar screen to type your question or comment at any time. NOW: Use the Chat Box to sign in. 1) Enter your organization and names of all people in the room. 2) Send to “HOST” 3) Click “SEND” Starting at the Beginning: Two Promises We Make to Our Patients We will do everything that we know will help. We will do nothing that will harm. 3 Copyright, The Reinertsen Group How well do we keep our promises? • Help: –55% –10-1 • Harm: –1% –14.6% –200,000 + 4 How Hazardous Is Health Care? (Leape) DANGEROUS (>1/1000) 100,000 REGULATED ULTRA-SAFE (<1/100K) HealthCare Total lives lost per year Driving 10,000 1,000 Scheduled Airlines 100 Mountain Climbing Bungee Jumping 10 Chemical Manufacturing Chartered Flights European Railroads Nuclear Power 1 1 10 100 1,000 10,000 100,000 Number of encounters for each fatality 1,000,000 10,000,000 Are we seeing all the harm? Inpatient Surgical Record Review of 854 patients in 11 US hospitals… • Found 14.6% of patients had a Surgical Adverse Event (SAE) • 44% of SAEs caused increase LOS or readmit • 8.7% required life-saving intervention or resulted in permanent harm or death • “…Most of the events identified by Trigger Tool review had not been detected or reported via any other existing mechanism.” 6 References • McGlynn EA, SM Asch, J Adams, J Keesey, J Hicks, A DeCristofaro, EA Kerr: The Quality of Health Care Delivered to Adults in the United States. New England Journal of Medicine 2003, 348: 2635-2645 • Zhan C, Miller MR. Excess length of stay, charges, and mortality attributable to medical injuries during hospitalization. JAMA. 2003;290:1868-1874 • Amalberti R, Y Auroy, D Berwick, P Barach. Five System Barriers to Achieving Ultrasafe Health Care Ann. Int. Med. 2005; 142: 756-764 • Griffin, FA and DC Classen. Detection of adverse events in surgical patients using the trigger tool approach. Qual. Saf. Health Care 2008; 17: 252-8 7 Reliability: Defect-free operation, over time, regardless of… 8 How Reliable is Health Care? (Amalberti, Nolan) Chaos Processes are largely custom-crafted each time 10-1 Standard specs, training, trying hard 10-2 Standard process; redundancy, habits and patterns… 10-3 , 10-4 10-5 HRO culture; Loss of Obsession with identity Each doctor writes individual orders, gives to RN 5 people describe 5 processes; feedback on compliance 5 people describe 1 process; multi-disc. rounds External approval necessary for certain orders Equivalent actor Preventing, treating acute and chronic disease in US Surgical checklists and harm Best hospitals Core Measures ADEs per 1000 doses, blood banking Safety in anesthesia failure, deference to expertise… Let’s talk about the catastrophic processes—the right side of this table. Could we do better than this? 10 What is possible in safety? Commercial Aviation 11 Scrams per 7,000 hours Unplanned Automatic SCRAM Rate – US Nuclear Power Plants Year The unplanned automatic scrams per 7,000 hrs critical indicator tracks the median scram (automatic shutdown) rate for approximately one year (7,000 hrs) of operation. Unplanned automatic scrams result in thermal and hydraulic transients that affect plant systems. The scram rate has been significantly reduced since 1980. In 2000, 59% of operating plants had zero automatic scrams. Source: Statistics Show US Nuclear Power Plants Always Improving, Nuclear News, May 2001 12 Class A Mishaps/100,000 Flight Hours Naval Aviation Mishap Rate 60 776 aircraft destroyed in 1954 Angled Carrier Decks Naval Aviation Safety Center 50 NAMP est. 1959 15 aircraft destroyed in 2008 RAG concept initiated 40 NATOPS initiated 1961 Squadron Safety program 30 System Safety Designated Aircraft 20 ACT 10 HFC’s 1.64 0 50 65 Fiscal Year 80 08 Source: www.safetycenter.navy/mil ORM Flight Mishap Rate 13 US Nuclear Powered Submarines 5,500 cumulative years of nuclear reactor ops 127 million miles submerged (264 round trips to moon) Operated by 20 year olds Zero reactor accidents 14 Highly Reliable Organizations (HROs) “operate under very trying conditions all the time and yet manage to have fewer than their fair share of accidents.” 15 Copyright 2006 Healthcare Performance Improvement, LLC. ALL RIGHTS RESERVED. How do they do that? 16 Culture and Structure: hierarchy, transparency, safety rules, accountability… Technology and Environment: error proofing, distractions… Processes: workflow scheduling, complexity… Harm Human error (slip, lapse, reliance on memory, confirmation bias…) The Swiss Cheese Model 17 Adapted from James Reason Managing the Risk of Organizational Accidents 1997 Culture: error has Nurse doesn’t check happened before; no dose closely in rush to one questions 1200 mg get all QD doses in at 8 in IV bag rather than am irrigation Rx Computer system accepts “1200 mg IV” Renal failure, death Pharmacist clicks 1200 rather than 120 mg on computer picklist 1200 mg Tobramycin IV 18 Errors, Harm, Negligence, and Intent Errors Skill based Rule based Knowledge based Harm Negligence, Reckless Disregard 19 Intent 2. Work as leaders to detect and plug the holes in processes, structures, cultures, and technologies Harm 1. Accept that human error will occur. Use human factors engineering to reduce the likelihood of it, and develop a culture of accountability Two Key Approaches to Higher Levels of Reliability for “Immediately Catastrophic” Processes 20 Working both the Sharp End and the Blunt End #1 Reduce likelihood of individual behavior failures •Competency •Consciousness •Communication •Compliance •Critical thinking #2 Find and Fix system “latent errors” •Structure •Culture •Policy/protocol •Process •Technology & environment Adapted from Kerry Johnson, HPIpyance. 21 Reducing the likelihood of skill-based (automatic) errors ─Fatigue: work schedules ─Distractions: “do not disturb” ─Error-proofing, design: Separating IV and topical dose forms in computer ─Visual signals Stop, Think, Act and Review 22 Reducing the likelihood of rules-based errors • Make it easy to comply with the rule • Increase the perception of “likelihood of being observed” while carrying out the rule • Increase the perception of the risk of non-compliance 23 Establish Accountability Accountability from Self Accountability from Peers Optimal Accountability Accountability from Leaders Healthcare Performance Improvement 24 How Reliable is Health Care? (Amalberti, Nolan) Chaos Processes are largely custom-crafted each time 10-1 Standard specs, training, trying hard 10-2 Standard process; redundancy, habits and patterns… Each doctor writes individual orders, gives to RN 5 people describe 5 processes; feedback on compliance 5 people describe 1 process; multi-disc. rounds Preventing, treating acute and chronic disease in US Surgical checklists and harm Best hospitals Core Measures 10-3 , 10-4 10-5 HRO culture; Loss of Obsession with identity failure, deference to expertise… Let’s look at the left External Equivalent side ofactor this approval necessary table for certain orders ADEs per 1000 doses, blood banking Safety in anesthesia 25 Discussion for NAPH Leaders • Describe the reliability of a safety process your teams are currently working to improve. Where does that process fall on the reliability grid? • What are your current ideas for making the process more reliable? • What is your aim: how reliable are you trying to become? 26 Why do we get “stuck” at low levels of reliability? • We tend to rely on vigilance and hard work • We focus on outcomes rather than process • We fail to design and implement standard work • We don’t understand and use sophisticated designs for reliability 27 Improvement Concepts Associated with 10-1 Performance • • • • • • • Common equipment Standard order sets Care protocols and pathways Written policies/procedures Personal check lists Feedback of information on compliance Suggestions to work harder, pay closer attention… next time 28 • Awareness and training Improvement Concepts Associated with 10-2 Performance • Build decision aids and reminders into the system • Make the desired action the default • Redundancy • Scheduling • Take advantage of existing habits and patterns of work • Standardize who, where, when…not just what (Standard work, not standard specs) 29 Examples of Level 2 Concepts for CHF or CAP • Decision aids or reminders in real time: – Standing order set is placed on front of chart when decision to admit is made • Desired action the default: – All patients with diagnosis of pneumonia will get pneumovax by nurse, with or without specific order • Redundancy: – Multidisciplinary rounds on every patient daily – Home visit to check on meds after 48 hours 30 Examples of Level 2 Concepts for CHF or CAP (2) • Scheduling: – Make a discharge appointment for all patients at least 24 hours prior to discharge • Smooths nursing workflow • Starts process of discharge instructions well in advance • Engages family members in planning • Take advantage of existing habits and patterns – Visual cue to “start pre-op antibiotics” based 31 on measured flow of pre-op work Example of Level 2 Concepts • Bundles –clusters of evidence based services in space and time, treated as “all or none” e.g. “sterile technique in the OR”, ventilator bundle… 32 Key Learning Points for Leaders • Hard work and vigilance alone will condemn the team to 10-1 performance at best • If 10-2 change concepts do not make up at least 25% of the improvement effort on a given project require the team to rethink the design 33 Why is health care so unreliable? • We tend to rely on vigilance and hard work • Greater focus on outcomes than process • We fail to design and implement standard work • We don’t understand and use sophisticated designs for reliability 34 Outcomes versus Process • Biology protects us: each process defect doesn’t necessarily lead to a bad outcome e.g. hand-washing • Systems protect us…except when they don’t e.g. 1200 mg tobramycin i.v. • Benchmarks reassure us…e.g. “2.0 BSI’s per 1000 line hours is better than the benchmark 35 so our processes must be OK” Why is health care so unreliable? • We tend to rely on vigilance and hard work • We focus on outcomes rather than process • We fail to design and implement standard work • We don’t understand and use sophisticated designs for reliability 36 Level 2 Concept: Standard Work Is this what you mean by standardization? • Months of meetings designing a care pathway, standing order set, protocol… • Focus is on WHAT should be done (not HOW). • 10,000 copies of the final version are printed up. • Changes to the final version are discouraged. • Physicians are encouraged to “opt in”. • Even though the protocol or order set has never been tested in the field 37 Design Design Design Design Approve Conference Rooms Real World Implement 38 A Better Way to Standardize • Spend no more than one meeting on “what”. • Start testing one way to do the “what” in the field, on a small scale. • Encourage many rapid tests of change in how, when, where, who…and in the what, if necessary…to make the standard way work well for 90% of doctors and nurses. • Once you’ve got it right, expect all the doctors and nurses to use it (opt out if you have to, not “please opt in”). 39 Conference Rooms Approve (if necessary) Design Test and Modify Test and Modify Test and Modify Implement Real World 40 “Practice the science of medicine as a team, and the art of medicine as individuals.” 41 OK…But what do Leaders do? • • • • Choose the outcomes that you want to achieve Hypothesize: process and outcome Set reasonable timelines Expect teams to – Achieve process reliability at 95% – Use good design principles, not just vigilance and hard work – Test designs frequently, on small scale, not in one big spasm • If the process gets to 95% and the outcome doesn’t improve… – Check the data on the process – Revise your theory 42 A Key Question Leaders Should Ask Is there a logical, evidence-based connection between the process the team is trying to improve, and the outcome you wish to achieve? 43 Is this a good plan to reduce injuries from falls? Outcome Goal Decrease Falls 44 Key Drivers Processes Reliable risk Red booties assessment for at risk of patients for patients falls The Fact There is no good medical evidence that risk assessment actually reduces falls! 45 Evidence-Based Thinking! Outcome Goals Decrease Falls 46 Key Drivers Toileting in at risk patients Processes Every 2 hour toileting rounds on at risk patients Summary: To become more reliable for non-catastrophic processes… • • • Choose a process that has a high likelihood of affecting the outcome of interest Set an aim for 10-2 reliability of the process Move toward that aim in three steps: 1. Segment and standardize to get the basic process to solid 10-1 using Level I and Level II concepts 2. Identify remaining defects and mitigate them in real time to get to 10-2 3. Redesign the process to reduce likelihood of defects 47 Where to read more…. • http://www.ihi.org/IHI/Topics/Reliability/ 48 Next Month: Thursday, May 23 9am PT/10am MT/11am CT/ 12am ET When Things Go Really Wrong: Responding to Organizational Crises With special guest speaker Jim Conway! Assignment: • Find out if your organization has a crisis management plan for safety disasters. Then ask: • What is the plan for notifying the board? Which board members? When? By whom? • Who will speak for the organization? Who is to speak to family members? • What training is in place? Be prepared to discuss your plans on the webinar. 49 SAVE THE DATE! Leadership for Safety: Yes, It’s Personal A Workshop for Boards, C-Suite, and Senior Leaders June 19, 2013 8:00am – 5:00pm Westin Diplomat Resort in Hollywood, FL 50 THANK YOU FOR JOINING US Feedback survey can be accessed in chat box. 51