Tricks of the trade: delivering reliable healthcare

To our patients, the notion that we consistently deliver only a proportion of the care that we intend would probably come as something of a surprise. We would all be disappointed if our garage mechanic chose to complete only some of the elements required for a routine car service (and charged the standard fee), and yet we routinely tolerate such performance in our day-to-day clinical practice. Two concrete examples from the UK are the implementation of enhanced recovery through the national Enhanced Recovery Partnership Program (ERPP) 1 and the provision of emergency general surgical care documented through the National Emergency Laparotomy Audit (NELA) 2. In both cases, striking variations in clinical practice were evident between institutions across England and improved adherence to accepted standards of care was associated with better outcomes for patients. Variation within healthcare may, of course, be appropriate; for example, differences in prescription rates of drugs to treat malaria between tropical and non-tropical countries. However, more important is the concept of unwarranted variation, originally popularised by Jack Wennberg in relation to the variations in healthcare associated with geographical location, where variation is not accounted for by ‘type or severity of illness or by patient preferences’ 3, but is dependent on the performance of individuals, teams or institutions. Such variation is both common and persistent. How is it that there are such entrenched, interinstitutional variations in care? And why is so little energy focused on minimising such unwarranted variation in our daily practice? Such challenges have been met, and overcome, in other walks of life. In the commercial sector, US-based business thinkers, including Juran and Edwards Deming 4, developed the techniques of quality improvement. This, in turn, led to business philosophies focusing explicitly on ensuring high reliability, such as ‘six sigma’ and Total Quality Management 5. Effective operations based on consistently reliable high-quality systems are the lifeblood of a successful company. But, unfortunately, it is not clear that the same can be said with confidence of healthcare providers. Beacons of hope shine: Don Berwick's Institute for Healthcare Improvement (IHI) has been an effective champion of improving reliability within healthcare 6. The 2004 IHI publication ‘Improving the reliability of healthcare’ 7 defines reliability as ‘failure-free operation over time’, to which a numerical value can be attached via the equation: Reliablity = number of actions that achieve the intended result ÷ total number actions taken From this equation, an index of unreliability (failure rate) can be defined by 1-Reliability and is commonly described as an exponent of base 10. Thus, 10−1 represents one failure in 10 events, 10−2 represents one failure in 100 events and 10−6 represents one failure in 1,000,000 events; aviation passenger safety and nuclear power plant design briefs operate to this criterion 7. In this issue of Anaesthesia, Smirk et al. report the use of a ‘Greenie board’ as a tool to promote adherence to their established local enhanced recovery after surgery protocol 8. The Greenie board is a simple tool based on a visual feedback system used by the US Navy to monitor pilot performance of end-of-mission landings onto aircraft carrier decks. The authors adapted this system to benchmark adherence to seven measures describing five elements of an established enhanced recovery protocol. The level of individual clinician adherence to these measures was recorded on the Greenie board, colour-coded with green for ≥ 6 out of 7 measures (hence ‘Greenie board’), yellow for 5, orange for 4 and red for < 4 measures per patient. Using a before-and-after study design within one institution, they showed that implementation of the Greenie board resulted in improvements in adherence to five out of seven elements: antibiotic prophylaxis; antibiotic re-dosing; goal-directed fluid therapy; anti-emetic prescription; and postoperative intravenous fluid rate. They also showed an increase in the overall number of ‘green’ cases, which more than doubled from 33% to 72%. The two elements that did not show a significant improvement (use of a forced air warming blanket and achievement of temperature ≥ 36°C on admission to the recovery area) both increased, albeit by a modest amount and from a high baseline (>70%). Although postoperative outcomes including complications and length of hospital stay were not improved, there was a substantial reduction in postoperative nausea and vomiting (40.0% to 23.1%) that may, at least in part, be attributable to the improved administration of prophylaxis (49% to 70%). It is notable that this study documents improvements in reliability that correspond to failure rates in the 10−1 to 10−2 domains – several orders of magnitude lower than commonly accepted standards in other safety-focused industries. Although on occasion there may be clinically-relevant reasons for non-adherence to planned care, this is not typically the case. Justifiable non-adherence should be clearly documented through exception reporting encompassed within an adherence reporting system. This study clearly demonstrates that a low-cost and simple innovation can improve the reliability of delivery of planned care. Failure to ‘shift the needle’ in terms of reduced length of stay may be related to the fact the study focused on a subset of enhanced recovery elements that were exclusively within the control of the operating room anaesthetist or that the outcomes measured did not encompass patient-reported measures that may have been impacted by this intervention 9. It is interesting to speculate whether a broader scope of the Greenie board approach, involving the whole multidisciplinary team and all elements of enhanced recovery, might have resulted in reduced morbidity and hospital stay. Improved compliance with enhanced recovery protocols is, in general, associated with improved clinical outcomes in cross-sectional observational 1, retrospective 10 and prospective evaluations 11. So, what are the obstacles to such approaches in the healthcare setting? One obstacle may be the persistent meme of the physician as artist, using his/her skill and judgement to tailor an individualised package of care for each patient 12. Such thinking has been embedded in the training culture of the so-called craft specialties, including anaesthesia and surgery, for many years. The idea that the consistent reliable application of evidence-based best practice might be better for patients is, at least in part, antithetical to such a notion. However, the valid belief that strict adherence to evidence-based practice may not always be in a patient's best interest 13 is misused when interpreted as a license to ignore clinical recommendations without clear justification. Another obstacle may be resistance to ‘league tables’ in healthcare. Comparative ranking of providers by outcome performance has been heavily criticised for a variety of reasons, including: stigmatisation of poor performers 14; overinvestigation of unusual performance 15, 16; and an overall negative impact on public trust and professional moral 17. In particular, simple ranking fails to take account of differences in case-mix or the role of chance 15, 18. For outcome measures, risk-adjusted control charts or funnel plots offer more sophisticated and valid approaches under such circumstances. In contrast, league tables of evidence-based process measures can be an effective driver of reliable healthcare delivery 8, 19 and are unlikely to be associated with patient harm. We know that the provision of data alone is insufficient to change practice 20. Simple tools such as the Greenie board offer an effective means to nudge clinicians towards ‘better’ behaviour through the catalyst of competition – the very essence of a league table. Such techniques have enormous potential to drive improvements in care alongside more sophisticated quality improvement techniques including control charts and other mechanisms of data feedback 21. Looking forward, the application of new technologies to rapidly and effectively collect and communicate such data to clinicians, occurring as near to real-time as possible, will be facilitated by the ‘Internet of things’ 22. So, what are the take-home messages from this study? First, a low-cost simple intervention, borrowed from an environment where high-quality performance is literally ‘mission critical’, was effective at improving the reliability of delivery of anaesthetic care. Second, that competition can be a force for good in healthcare. The use of league tables, or similar interventions such as the Greenie board, for driving improvements in adherence to process measures, in contrast to outcome measures, is effective and unlikely to result in harm. Finally, this study should focus us on the critical importance of reliability of healthcare delivery. Although incremental improvements in the performance of healthcare processes may seem mundane, the consistent delivery of best practice to every patient in every institution, every day, may achieve more for patients than many randomised, controlled trials could ever hope to. MG serves on the Board of ERAS UK, Oxygen Control Systems Ltd, the medical advisory board of Sphere Medical Ltd and the international advisory board of the American Society of Enhanced Recovery, MG has received honoraria for speaking and/or travel expenses from Edwards Lifesciences, Fresenius-Kabi, BOC Medical (Linde Group), Ely-Lilly Critical Care, and Cortex GmBH. MG is Executive Chair of the Xtreme-Everest Oxygen Research Consortium.