High Reliability healthcare David Mudd

High reliability healthcare: how much choice can we afford?: 

High reliability healthcare: how much choice can we afford? David Mudd Head of Modernisation North West London SHA 4 August 2004 david.mudd@nwlha.nhs.uk

In this session you can: 

In this session you can see comparisons of reliability in healthcare and other industries; look at how to design for better reliability; look at the benefits and costs of high reliability organisations; think about how you can improve the reliability of your service

The basics: 

The basics Joined-up services - as seen by users Shared information Choice – providers, settings, date and time Redesigned, streamlined services – fewer visits, no avoidable waits Better working lives for NHS staff

Higher expectations: 

Higher expectations over time, the basics get taken for granted then we expect high reliability examples – cars, TVs, food industry, civil aviation, nuclear industry

High reliability healthcareDon Berwick, Tom Nolan: 

High reliability healthcare Don Berwick, Tom Nolan Reliability = number or % of actions that achieve the intended results “Getting it right, every time”

Healthcare process reliabilities: 

Healthcare process reliabilities

Reliability in healthcare and other industries (Amalberti; Espinosa & Nolan): 

Reliability in healthcare and other industries (Amalberti; Espinosa & Nolan) Himalaya mountaineering Fatal iatrogenic adverse events Cardiac surgery (Patient ASA 3-5) Medical risk (total) Road safety Chemical industry (Europe) Civil aviation Railways - France Nuclear power industry Anaesthesia ASA1 Blood transfusion ED reading X-rays (97%) ED reading X-rays - improved (99.7%)

Design for Reliability (Berwick & Nolan): 

Design for Reliability (Berwick & Nolan) Level 1: Intent, vigilance and hard work Level 2: Design informed by reliability science and research in human factors Level 3: Design of high reliability organisations

Level 1: “Intent, vigilance, hard work”: 

Level 1: “Intent, vigilance, hard work” Standardized protocols, feedback, training, checklists Guidelines easy; getting people to follow them is difficult But: good evidence exists that this will get you to 10-2 (99%)

Group discussion (5 mins): 

Group discussion (5 mins) Recall an experience – in any setting – in which the request that you ‘try harder’, ‘be careful’, or ‘stay alert’ improved your performance. Why did that work? Identify a process in your workplace that relies on vigilance. What would you estimate its reliability to be?

Factors affecting human vigilance: 

Factors affecting human vigilance Fatigue Environmental conditions Task design Psychological conditions Competing demands

Level 2: Design informed by reliability science and research in human factors: 

Level 2: Design informed by reliability science and research in human factors Redesign the system; don’t rely on checking Decision aids and reminders built into the system; automation;evidence as the default; redundancy; scheduling; connection to habits Can take you from 10-2 (99%) to 10-3 (99.9%)

Three-level design of safe and reliable systems: Prevent – Identify - Mitigate (Norman): 

Three-level design of safe and reliable systems: Prevent – Identify - Mitigate (Norman) Design the system to prevent failure Design the system to make failures visible if they do occur Design procedures to mitigate the harm caused if failures are not detected and intercepted

Physical design, human behaviour (Norman): 

Physical design, human behaviour (Norman) Constraints physical (eg square peg round hole) cultural (eg Red for Stop) Memory eg numbers, passwords, dates Mapping eg my cooker

Natural mapping – my cooker (Norman): 

Natural mapping – my cooker (Norman)

Advantages of automation: 

Advantages of automation Increased capacity and productivity Reduction of manual workload and fatigue Relief from routine operations Relief from small errors More precise handling of routine operations Economical utilisation of machines Damping of small differences

Disadvantages of automation: 

Disadvantages of automation Seen as dehumanising – lower job satisfaction Low alertness of human operators Systems are fault intolerant – may lead to larger errors Silent failures Lower proficiency of human operators Over-reliance – complacency, uncritical acceptance of results False alarms

Computers are incredibly fast, accurate and stupid; humans are incredibly slow, inaccurate and brilliant; together they are powerful beyond imagination: 

Computers are incredibly fast, accurate and stupid; humans are incredibly slow, inaccurate and brilliant; together they are powerful beyond imagination Albert Einstein

What’s the contribution of IM&T in healthcare?: 

What’s the contribution of IM&T in healthcare? Shared electronic record – one set of data, available to all Protocol-driven care as the default – standardising on best practice Decision aids Scheduling Making errors visible and retrievable Letting every player see the whole system

Level 3: Design of high reliability organisations: “ultra-safe” (Weick; Amalberti): 

Level 3: Design of high reliability organisations: “ultra-safe” (Weick; Amalberti) Accepting limits on discretionary actions (eg pilots told when it’s safe to take off/land) max safety up to 10-2 Abandoning autonomy (eg drive on the left) max safety up to 10-3 Moving from ‘craftsmanship’ to ‘equivalent actor’ (eg anaesthetics v surgery) max safety up to 10-4 Sharing the residual risk – how the system deals with failure (eg KAL 007) max safety up to 10-5 Managing the invisibility of risk (“we’ve never seen one of these before” max safety up to 10-6

Some outcomes of focus for a high reliability healthcare organisation(Berwick): 

Some outcomes of focus for a high reliability healthcare organisation (Berwick) No needless deaths No needless pain No feelings of helplessness No unnecessary waiting No waste

Constraints : 

Constraints Constraints on patient choice Constraints on professional autonomy Constraints on the factors which distinguish healthcare organisations (competitive advantage)

Challenges : 

Challenges Buy-in from clinicians and patients Protocol-driven care as the default – standardising on best practice Sharing information – with whom? Decision aids or decision-taking? Scheduling – are we ready for it? Making errors visible and retrievable – aligning the incentives and penalties

Summary : 

Summary Reliability in healthcare is a long way behind other safety-critical industries We can improve reliability by better design equipment and systems, and by standardisation Beyond this, improvement comes with reduced choice – standardisation/doing it by the book Clinicians need to work with standardisation, but be prepared to respond with autonomy Sophisticated design of human interactions and working relationships

Table discussion: 

Table discussion Think about the clinical service you provide/manage: What changes must you make to get to 10-4 performance? What implications does this have for how you select and train your staff?

References/further reading: 

References/further reading The Design of Everyday Things. Donald A Norman; MIT Press; 1998 Handbook of Human Factors and Ergonomics. G Salvendy (ed); John Wiley, New York; 1997 Managing the Unexpected. Weick and Sutcliffe; Josey Bass; 2001 System changes to improve patient safety. Nolan TW. BrMedJ 2000;320: 771-773 (18 March) Reducing errors made by emergency physicians in interpreting radiographs: longitudinal study. Espinosa JA, Nolan TW. BrMedJ 2000; 320: 737-740 (18 March) The quality of care delivered to adults in the United States. McGlynn EA, Asch SM, et al. N Engl J Med, 348(26); 2635-2645 26 June 2003 Revisitjng safety and human factors paradigms to meet the safety challenges of ultra complex and safe systems. Amalberti R, 2001. In Challenges and pitfalls of safety interventions, Willpert B & Fahlbruch B. Elsevier, North Holland Wiener EL, Nagel DC, ed. Human Factors in Aviation. New York: Academic Press, 1988. Various presentations & papers by Rene Amalberti on the Internet (via Google)